The Vertebral Column

The first edition of Grieve's Modern Manual Therapy - The Vertebral Column was quickly recognized as a milestone in the field of non-surgical treatment of back problems. This third edition maintains the objectives of the original editor, Gregory Grieve, to bring together the latest state-of-the-art research, from both clinical practice and the related basic sciences, which is most relevant to practitioners. The new international editorial partnership of Jeffrey Boyling and Gwendolen Jull has ensured a new look to the third edition, with the inclusion of contributions on key and cutting-edge work from around the world. As in the two previous editions the topics addressed and the contributing authors have been selected to reflect the best and most clinically relevant contemporary work going on in the field. The text is grouped into five main sections:

Grieve's Modern Manual Therapy has been called 'the manual therapist's Bible'. This new edition will justify the continuing use of that term. No other text in the field presents such an international spread of up-to-date and cutting edge research related to the clinical practice of manual therapy in relation to the spine. The aim of the editors has been to create a real encyclopaedia of 'state-of-the-art' knowledge, which is current, comprehensive and accessible. In achieving their objective they have ensured that the book will continue to be used as a textbook by those wanting to become manual therapy practitioners, as well as by experienced therapists wanting to revise or update their knowledge. No-one who aspires to be a manual therapist can afford to be without their own copy of this text. Reviewers' comments on the First Edition
'The outstanding value of this book is that it brings together so many different viewpoints on manual therapy from many different countries. A most impressive book - of use to all manual therapists, however experienced ...'

Section 1 looks at the scope of manual therapy in the future. Section 2 covers the foundation sciences relating to manual therapy, principally anatomy, biochemistry, clinical biomechanics, motor control and the physiology of pain. Section 3 addresses advances in the clinical sciences relating specifically to manual therapy . of the spine. Section 4 deals with the clinical sciences and practices within manual therapy, such as specific therapeutic exercise, taping, clinical reasoning and pain management. Section 5 looks at the issues of establishing an evidence base for manual therapy.


'An impressive book by any standards. The material presented has both breadth and depth ... The volume is an invaluable source of reference to therapists treating musculoskeletal disorders.'



Physiotherapy Practice Reviewers' comments on the Second Edition
' ... this book is one of the few resources that contains such a voluminous amount of high-quality information.'


Key Features

Compiled and edited by two internationally recognized leaders in the field who are both actively involved in research and clinical practice Includes 43 chapters written by an invited team of 68 contributors from around the world, all of them recognized leaders in their specialist areas Covers problems and techniques affecting the management of conditions relating to all parts of the vertebral column Highly illustrated with 270 illustrations, both photographs and line drawings All chapters are based on published research, making the book truly evidence-based

'It is a comprehensive reference source of current thinking in this rapidly expanding speciality .. .'


Manipulative therapist
'With 54 contributors and 4454 references, this book will ensure that the manual therapist is kept awake and informed.'


All professionals involved in the assessment and treatment of spinal conditions will find in Grieve's Modern Manual Therapy an authoritative reference work which is essential to their practice.



This product is appropriate for:


• • • •

manual therapy physiotherapy chiropractic osteopathy


3 07 rl443� .71Ir 780 O. 553

Grievers Modern Manual Therapy

In memory of Gregory Peter Grieve, 77 December 1918 - 24 April 2001

For Churchill Livingstone: Editorial Director, Health Professions: Project Development Manager: Project Manager: Illustrations:

Mary Law Claire Wilson

Ailsa Laing PCA Creative Design: Judith Wright

Grievers Modern Manual Therapy
The Vertebral Column

Edited by

Jeffrey D. Boyl i ng

MSe (Land) BPhty (Hans) (Qld) GradDipAdvManipTher (SAlT) MAPA MCSP MErgS MMPA

Chartered Physiotherapist and Ergonomist, Hammersmith, London, UK

Gwendolen A. Jull MPhty GradDipManipTher PhD FACP
Professor and Head, Division of Physiotherapy, The University of Queensland, Australia

Foreword by

Professor Lance T. Twomey


Vice-Chancellor. Curtin University of Technology, Perth, Australia

/')\ �.A





An imprint of Elscvier Limited

e Longman Croup Limited 1986. 1994 e 2004, Elsevier limited. All rights reserved.
The right of Jeffrey 0 Bayling and Gwendolen A Jull 1'0 be identified as editors of this work has been .1sserh.'Ci by them in accordance with the Copyright, Designs and Patents Act 1988. No part of this publkation may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying. recording or othem'ise. without either the prior }X'rmission of the pUblishers or

licence permitting restricted copying in the United Kingdom ourt Road, London WIT 4LP. Permissions

issued by the Copyright Licensmg Agency. 90 Tottenham

may be sought directly from Elsevier's Health Sciences Rights Deparbnent in Philadelphia, USA: phone: (+1) 215 23S 7869, fax: (+1) 215 238 2239. ('-mail: You may also complete your request on·line viii the ELsevier Science homepage ( by !JCloolng' ustomer Support' and then 'Obtaining Permissions'. first edition 1986 Second edition 1994 Third wition 2004 ISBN ().I43 071551 British Library Cataloguing In Publication Data A catalogue record for thiS book is available from the British Library Library of Congress Cataloging in Publication Data A catalog record for this book is available from thc Ubrary of Congress lotc Every effort has been made by the Editors and the Publishers to ensure that the descriptions of the techniques included in this book are aCC\Jrate and in conformity with the descnptions published by their developers. The Publishers and the Editors do nol assume any responsibility for any injury and/or damage to persons or property arising out of or related to any usc of the material contained in this book. It is the responsibility of the treating practitioner, relying on indcpendent experience and knowledge of the patient, to determine the best treatment and method of application for the pallent, to make theIr own evaluation of their effectiveness and to check with the developers or teachers of the techniques they wish to use that they have understood them correctly.

TIle PublisJlt'r

your source tor books. journals and muttimecUa In the heotth sciences



Contributors Foreword xi

vii xiii xiv xv

11. The lumbar fasciae and segmental control P. J. Barker, C. A. Briggs


Preface to the third edition Preface to the second edition Preface to the first edition Acknowledgements xvi

12. Neurophysiology of pain and pain modulation A. Wright, M. Zusman 13. The effect of pain on motor control M. Galea 173


SECTION 1 Introduction to modern manual therapy
1. The future scope of manual therapy J. D. Boyling, G. A. Jull 3

14. The spine and the effect of ageing K. P. Singer


SECTION 3 Clinical sciences for manual therapy of the spine


SECTION 2 Foundation sciences for manual therapy
2. Comparative anatomy of the spinal disc S. Mercer 3. Comparative anatomy of the zygapophysial joints K. P. Singer, J. J. W. Boyle, P. Fazey 4. Kinematics of the spine S. Mercer 5. Chemistry of the intervertebral disc in relation to 31 17 9


15. How inflammation and minor nerve injury contribute to

pain in nerve root and peripheral neuropathies
J. Greening 16. Chronic pain and motor control G. L. Moseley, P. W Hodges 17. Cervical vertigo H. Heikkilii 18. The cervical spine and proprioception E. Kristjansson 243 233 215


functional requirements
J. P. Urban, S. Roberts


19. The vertebral artery and vertebrobasilar insufficiency D. A. Rivett


6. Clinical biomechanics of the thoracic spine including the



20. Mechanisms underlying pain and dysfunction in whiplash

S. J. Edmondston 7. Clinical biomechanics of the lumbar spine J. Cholewicki, S. P. Silfies 8. Clinical biomechanics of lifting S. Milanese 9. Motor control of the cervical spine E. A. Keshner 10. Motor control of the trunk P. W. Hodges 119 105 89 67

associated disorders: implications for physiotherapy management
275 M. Sterling, J. Treleaven, G. A. Jull 21. The cervical spine and headache G. A. Jull, K. R. Niere 22. 'Clinical instability' of the lumbar spine: its pathological 291

basis, diagnosis and conservative management
P. B. O'Sullivan 23. Abdominal pain of musculoskeletal origin V. Sparkes 333




24. Osteoporosis


35. Pelvic floor dysfunction in low back and sacroiliac

K. Bennell, J. Larsen



R. Sapsford, S. Kelley

SECTION 4 Clinical science and practices of manual therapy 365
25. Neurophysiological effects of spinal manual therapy T. Souvlis, B. Vicenzino, A. Wright 26. Manual therapy and tissue stiffness D. Shirley 27. Clinical reasoning in the diagnosis and management of 381 367

36. Vascular syndromes presenting as pain of spinal origin A. J. Taylor, R. Kerry 37. Adverse effects of cervical manipulative therapy D. A. Rivett 38. Managing chronic pain P. J. Watson 551 533


spinal pain


N. Christensen, M. Jones, I. Edwards 28. The integration of validity theory into clinical reasoning: a

SECTION 5 Establishing the evidence base for manual therapy 567
39. A case for evidence-based practice in manual therapy A. R. Gross, B. Chesworth, J. Binkley 40. Methodological and practical issues in clinical trials on 569

beneficial process?


A. M. Downing, D. G. Hunter 29. Management of mechanosensitivity of the nervous system

manual therapy


in spinal pain syndromes
T. M. Hall, R. L. Elvey


J. L. Hoving, G. A. Jull, B. Koes 41. Outcomes assessment and measurement in spinal

30. The use of taping for pain relief in the management of

musculoskeletal disorders


spinal pain
J. McConnell


R. A. H. M. Swinkels, R. A. B. Oostendorp 42. Critical appraisal of randomized trials, systematic reviews

31. The rationale for a motor control programme for the

of randomized trials and clinical practice guidelines
C. G. Maher, R. D. Herbert, A. M. Moseley, C. Sherrington, M. Elkins 451


treatment of spinal muscle dysfunction
C. A. Richardson, J. A. Hides


32. A therapeutic exercise approach for cervical disorders G. A. Jull, D. Falla, J. Treleaven, M. Sterling, S. O'Leary 33. A contemporary approach to manual therapy R. L. Elvey, P. B. O'Sullivan 34. The management of pelvic joint pain and dysfunction D. Lee, A. Vleeming 471

43. Establishing treatment guidelines for manual therapy of

spinal syndromes


A. R. Gross, L. Hurley, L. Brosseau, I. D. Groham



vi i


Priscilla J. Barker


Angela M. Downing


Department of Anatomy and Cell Biology, University of Melbourne, Victoria, Australia
Kim Bennell
BAppSc(Physio) PhD

Senior Lecturer, School of Allied Health Professions, Faculty of Health and Social Care, University of the West of England, Bristol, UK
Stephen J. Edmondston DipPT AdvDipPT(ManTher) PhD

Associate Professor, Centre for Health, Exercise and Sports Medicine, School of Physiotherapy, University of Melbourne, Victoria, Australia
Jill Binkley

Associate Professor of Musculoskeletal Physiotherapy, School of Physiotherapy, Curtin University of Technology, Perth, Australia
Ian Edwards
PhD GradDipPhysio(Ortho) MAPA

Assistant Professor (PT), McMaster University, Hamilton, Ontario, Canada; Director, Sentinel Associates, Alpharetta, Georgia, USA
Jeffrey J. W. Boyle
BSc(Phty) GradDipManipTher

Physiotherapist, The Brian Burdekin Clinic, and Lecturer, School of Health Sciences, University of South Australia, Adelaide, Australia
Mark Elkins

Lecturer, Centre for Musculoskeletal Studies, School of Surgery and Pathology, University of Western Australia, Australia
Jeffrey D. Boyling
MSc(Lond) BPhty(Hons)(Qld) GradDipAdvManip

Centre for Evidence-Based Physiotherapy, The University of Sydney, and Royal Prince Alfred Hospital, Sydney, Australia
Robert L. Elvey
BAppSe(Physio) GradDipManipTher


Chartered Physiotherapist and Ergonomist, Hammersmith, London, UK
Christopher A. Briggs
DipEd BSe MS PhD

Manipulative Physiotherapist, Senior Lecturer in Manipulative Physiotherapy, Curtin University of Technology, Perth, Western Australia
Deborah Falla
BPhty(Hons) PhD

Department of Anatomy and Cell Biology, University of Melbourne, Victoria, Australia
Lucie Brosseau

Research Officer, Department of Physiotherapy, University of Queensland, Brisbane, Australia
Peter Fazey
BAppSc(Physio) GradDipManipTher

Associate Professor, School of Rehabilitation Sciences, University of Ottawa, Ontario, Canada
Bert M. Chesworth

Lecturer, Centre for Musculoskeletal Studies, School of Surgery and Pathology, University of Western Australia; Private Practitioner, Perth, Australia
Mary Galea
BAppSc(Physio) BA GradDipPhysio(Neuro)

Research Director, Ontario Joint Replacement Registry, London Health Sciences Centre, London, Ontario, Canada
Jacek Cholewicki

Associate Professor, Biomechanics Research Laboratory, Department of Orthopaedics and Rehabilitation, Yale University School of Medicine, Connecticut, USA
Nicole Christensen

GradDipNeurosciences PhD

Professor of Clinical Physiotherapy, School of Physiotherapy, The University of Melbourne, Victoria, Australia
Ian D. Graham

Assistant Professor, Orthopaedic Curriculum Coordinator, Department of Physical Therapy, Mount St Mary's College, and Clinical Faculty, Kaiser Permanente Los Angeles Manual Therapy Fellowship, Los Angeles, USA

Associate Professor, School of Nursing, University of Ottawa; Senior Social Scientist, Associate Director, Clinical Epidemiology Program, Ottawa Health Research Institute; Associate Professor, Medicine and Epidemiology and

vi i i


Community Medicine, University of Ottawa, Canada; CIHR New Investigator
Jane Greening

Emily A. Keshner PT EdD

Senior Clinical Research Scientist, Sensory Motor Performance Program, Rehabilitation Institute of Chicago; Research Professor, Department of Physical Medicine and Rehabilitation, Feinberg School of Medicine, Northwestern University, Chicago, USA
Bart Koes

Consultant Physiotherapist, Dartford, Gravesend and Swanley Primary Care Trust, NHS Kent; Senior Honorary Research Fellow, Physiology Department, University College London; Research Fellow, London South Bank University, UK
Anita Gross
MSc BHScPT GradDipMarupTher FCAMT

Professor of General Practice, ErasmusMC, University Medical Center, Rotterdam, The Netherlands
Eythor Kristjansson PT PhD ManipTher BSc

Associate Clinical Professor, School of Rehabilitation Sciences, McMaster University, Hamilton, Ontario, Canada
Toby M. Hall
MSc GradDipManipTher

Private Practitioner, Reykjavik, Iceland
Judy Larsen

Manipulative Physiotherapist, Adjunct Senior Teaching Fellow, School of Physiotherapy, Curtin University of Technology; Director, Manual Concepts, Perth, Australia
Hannu Heikkila

Physiotherapist and private practitioner, Wesle1j Hydrotherapy Centre and St Andrew's Hydrotherapy Centre, Brisbane, Queensland, Australia
Diane Lee

Specialist in Family Medicine and Physical Medicine & Rehabilitation, Department of Otorhinolaryngology, Northern Sweden University Hospital, Umea, Sweden
Robert D. Herbert
PhD BAppSc MAppSc(ExSpSc)

Education and Clinical Consultant, Ocean Pointe Physiotherapy Consultants, White Rock, British Columbia, Canada
Jenny McConnell
BAppSci(Phty) GradDipMarupTher MBiomedE

School of Physiotherapy, Faculty of Health Sciences, University of Sydney, Sydney, Australia
Julie A. Hides,
BPhty MPhtySt PhD

Director, McConnell and Clements Physiotherapy, Mosman, Australia
Christopher G. Maher
PhD GradDipAppSc BAppSc

Senior Lecturer, Department of Physiotherapy, The University of Queensland, Brisbane, Australia
Paul w. Hodges
PhD MedDr BPhty(Hons)

Associate Professor, School of Physiotherapy, Faculty of Health Sciences, The University of Sydney, Australia
Susan Mercer
BPhty(Hons) MSc PhD FNZCP

Professor and NHMRC Senior Research Fellow, Department of Physiotherapy, The University of Queensland, Brisbane, Australia

Senior Lecturer, Department of Anatomy and Structural Biology, University of Otago, Dunedin, New Zealand
Steve Milanese
(Ergonomics) BAppScGrad Cert(Sports Physiotherapy) MAppSc GradDip

J an Lucas Hoving



Senior Research Fellow, Department of Epidemiology and Preventive Medicine, Monash University, Melbourne, Australia
D. Glenn Hunter

Ergonomist, Rankin Occupational Safety and Health, Mile End, South Australia; Senior Research Officer, Centre for Allied Health Research, University of South Australia, Adelaide, Australia; Clinical Specialist - Musculoskeletal, St Mary's Hospital, London, UK
Anne M. Moseley PhD GradDipAppSc

Principal Lecturer, School of Allied Health Professionals, Faculty of Health and Social Care, University of the West of England, Bristol, UK
Laurie Hurley


Lecturer, Department of Physical Therapy, University of Toronto, Ontario, Canada
Mark Jones

Rehabilitation Studies Unit, The University of Sydney, Australia
G. Lorimer Moseley
PhD BAppSc(Phty)(Hons)

PT GradDipAdvanMarupTher MAppSc

Senior Lecturer, Director, Master of Musculoskeletal and Sports Physiotherapy, School of Health Sciences, Physiotherapy Discipline, University of South Australia, Adelaide, Australia
Gwendolen A. Jull
MPhty GradDipMarupTher PhD FACP

NHMRC Clinical Research Fellow, Senior Lecturer, School of Physiotherapy, The University of Sydney, Australia
Kenneth R. Niere
BAppSc(Physio) GradDipMarupTher MMarupPhysio

Professor and Head, Division of Physiotherapy, The University of Queensland, Brisbane, Australia
Susannah Kelley
BPhty MPhtySt

Lecturer, School of Physiotherapy, La Trobe University, Victoria, Australia
Shaun O'Leary
BPhty(Hons) MPhtySt

Musculoskeletal Physiotherapist, Performance Rehab, Brisbane, Australia
Roger Kerry

Department of Physiotherapy, University of Queensland, Brisbane, Australia
Peter B. O'Sullivan
DipPhysio GradDipMarupTher PhD

Lecturer, Division of Physiotherapy Education, University of Nottingham, Nottingham, UK

Senior Lecturer, Manipulative Physiotherapist, School of Physiotherapy, Curtin University of Technology, Perth, Australia



Rob N. B. Oostendorp


Valerie Sparkes


Professor in Allied Health Care, Centre for Quality of Care Research, University Medical Centre, Catholic University of Nijmegen, Nijmegen; Research Director, Dutch Institute of Allied Health Care, Amersfoort, Netherlands
Carolyn A. Richardson
BPhty(Hons) PhD

Lecturer, Department of Physiotherapy Education, University of Wales College of Medicine, Cardiff, Wales, UK
Michele Sterling BPhty GradDipManipTher MPhty PhD

Lecturer, Division of Physiotherapy, The University of Queensland, Brisbane, Australia
Raymond A. H. M Swinkels

Associate Professor and Reader, Department of Physiotherapy, University of Queensland, Brisbane, Australia
Darren A. Rivett
(ManipPhty) PhD BAppSe(Phty) GradDipManipTher MAppSe

Medical Centre Coevering, Geldrop; Manual Therapy, Faculty of Medicine and Pharmacology, Free University, Brussels, Belgium; Lecturer, University of Genoa, Italy; Lecturer, MSc Physical Therapy, Breda, The Netherlands
Alan J. Taylor

Associate Professor, Discipline of Physiotherapy, Faculty of Health, University of Newcastle, Australia
Sally Roberts

Physiotherapy Manager, Nottingham Nuffield Hospital, Nottingham, UK
Julia Treleaven

Director of Spinal Research, Centre for Spinal Studies, Robert Jones and Agnes Hunt Orthopaedic Hospital NHS Trust, Oswestry, and Reader, Institute of Science and Technol�gy in Medicine, Faculty of Health, Keele University, UK
Ruth Sapsford
AVA DipPhty

Division of Physiotherapy, The University of Queensland, Brisbane, Australia
Jocelyn P. Urban

Pelvic Floor Physiotherapist, Mater Misericordiae Hospital, Brisbane, Australia
Catherine Sherrington
BAppSe(Physio) MPH PhD

Physiology Laboratory, Oxford University, Oxford, UK
Bill Vicenzino PhD MSc
BPhty GradDipSportsPhty

Research Fellow, Prince of Wales Medical Research Institute, University of New South Wales, Sydney, Australia
Debra Shirley
(Cumb) PhD(USYD) BSe(UNSW) GradDipPhty(Cumb) GradDipManipTher

Senior Lecturer and Director, Musculoskeletal Pain and Injury Research Unit, Department of Physiotherapy, The University of Queensland, Brisbane, Australia
Andry Vleeming

Lecturer, School of Physiotherapy, Faculty of Health Sciences, The University of Sydney, Australia
Sheri P. Silfies

Chairman of the Advisory Board for the Spine and Joint Centre, Rotterdam, Netherlands
P aul J. Watson
PhD MSe BSe(Hons) DipPT MCSPg


Assistant Professor, Department of Rehabilitation Sciences, Drexel University, Philadelphia, USA
Kevin P. Singer

Senior Lecturer, Department of Health Sciences, University of Leicester, UK
Anthony Wright BSe(Hons) GradCertEduc MPhtySt PhD MMPA

Associate Professor and Head, Centre for Musculoskeletal Studies, School of Surgery and Pathology, The University of Western Australia, Perth, Australia
Tina Souvlis
BPhty(Hons) PhD

Professor and Head of School, School of Physiotherapy, Curtin University of Technology, Perth, Australia
Max Zusman
DipPhysio GradDipI-flthSe MAppSe

Lecturer, School of Physiotherapy, Curtin University of Technology, Perth, Australia

Lecturer, Division of Physiotherapy, The University of Queensland, Brisbane, Australia




Modern Manual Therapy of the Vertebral Column had a major impact when it was first published in 1986. It was a huge book,almost 900 pages long,containing scholarly and clin­ ical information from important international practitioners within or associated with the rapidly evolving discipline of manual therapy. The second edition (1994) was similarly large with two-thirds of the chapters containing new mate­ rial, while the remainder was substantially revised and updated. The third edition is entirely new and truly demon­ strates not only the evolution in the thinking and practice within this discipline,but also highlights a 'changing of the guard' as the early eminent authorities properly give way to younger scholars and clinicians. Along the way, this ensures that manual therapy continues to forge ahead into the 21st century with the vigour and vitality which were a hallmark of its beginnings. Only 11 of the 52 authors involved in the second edition have contributed to the third,and all of these have presented different topics to before. It is particularly sad to note that three major authors and world figures that presented their seminal work in the first two editions are now deceased. They were Greg Grieve from the United Kingdom, David Lamb from Canada and Brian Edwards from Australia. All three were charismatic leaders and educators, pre-eminent in their field, enthusiastic in their promotion of the disci­ pline and truly wonderful men. As an international com­ munity,we are much the poorer for their loss. Nevertheless, their legacy is demonstrated in the continued growth of and regard for manual therapy worldwide, which is well demonstrated by this third volume. Many of the new wave of authors have studied and worked with Greg,David and

Brian,each of whom would have been delighted to see their life's work so well amplified and extended. Manual therapists are problem solvers. Each patient presents a unique occasion for therapists to use their under­ standing of science and behaviour to work toward the sat­ isfactory resolution of spinal problems. W hile this volume provides an up-to-date account of the clinical skills and practices available to therapists,it does so in the context of science and evidence-based practice. It is in these latter areas that knowledge has expanded so dramatically in recent years. Science now provides a much more complete knowledge of the structure,function,movement behaviour and pathology of the vertebral column than it ever did in the past. At the same time,there is a greater understanding of the physiology and manifestation of pain from vertebral structures and the behaviour of people affected by spinal pain and movement disorders. It is this reliance on science and evidence-based practice that so distinguishes the man­ ual therapy of today from that of the mid-20th century. In developing this third edition, the editors have not made the mistake of staying with the tried and trusted for­ mat of the past. This is a bold book. It moves the discipline forward and, although it pays due respect to the past, it proudly strides into the future with new authors,good sci­ ence, great ideas and soundly based practice. I suspect that Gregory Grieve would have loved the ways in which his passion for the discipline and practice of manual therapy have been made manifest in this third edition.





Preface to the third ed ition '

Since the second edition of this book was published the world has changed. Only future generations will be able to judge whether it was in general for the better or for the worse. However, in the world of manual therapy the changes that have taken place have been for the better. This third edition comes some 9 years after the second and 17 years after the first. Some readers may consider the gaps between the three editions to be long but in reality change does not take place overnight. The pauses reflect the time taken for further maturity to occur within the field of manual therapy. Research that was being considered at the time of the second edition has now been undertaken and the results considered. Readers of this new edition will be able to benefit from that research. At the same time, how­ ever,previous editions are not obsolete but remain a valu­ able reference tool and, with the passage of time, will provide a useful barometer of how the focus of the profes­ sion has changed and matured. Churchill Livingstone were the publishers of the first and second editions of Modern Manual Therapy. In the inter­ vening period the Churchill Livingstone imprint first

became part of Harcourt Health Sciences and then, more recently, part of Elsevier Limited. Fortunately, the same team has been able to assist the editors to compile this edi­ tion. All the chapters are new and the line-up of authors has been changed to reflect retirements as well as new aspiring manual therapists at the forefront of research and practice. Sadly, Greg Grieve is no longer with us to share in the publication of this edition. However, his quest for knowl­ edge and for answers to questions lives on. On reflection it is clear to see that his thirst for knowledge was the forerunner of evidence-based practice. His publications in the field of manual therapy are proof of this. However, his attention to clinical detail should not be overlooked since it reinforced the reality of practice-based evidence. It is to be hoped that the reader will find a balance between evidence-based prac­ tice and practice-based evidence as they appear in this edi­ tion of Grieve's Modern Manual Therapy: The Vertebral Column.

London and Brisbane,2005

J. D. B. G. A. J.


Preface to the second ed ition

The retirement of Gregory Grieve left Churchill Livingstone with a superb text to be continued as well as with the task of finding a replacement editor. The fact that the second edition has been a joint effort is a reflection on the immense contribution to physiotherapy, and manual therapy in par­ ticular, that Gregory Grieve has made. The first edition reflected the leading edge of practice in the early 1980s, and it is to be hoped that this edition reflects the views of manual therapists in the early 1990s. This text is by no means meant to be exhaustive or repre­ sentative of the full spectrum of work being undertaken. That task represents a dream of past and present editors. The challenge to validate work has been taken up and it is reflected in the research work included in this text, as well as in the change of emphasis on examination as shown by the appropriate chapters. It is also pleasing to see new material developed by physiotherapists being added to the knowledge base. It is fitting that this new edition of Modern Manual Therapy is being published in the centenary year of the old­ est physiotherapy association, the Chartered Society of Physiotherapy. The very roots of the profession are steeped in manual therapy, and it is pleasing that one of the core skills is still at the heart of physiotherapy practice. It is almost 10 years since the first edition, which is still regarded as the standard text in the subject area, was pub­ lished. Consequently, the second edition is completely new,

with the inclusion of representatives of a new generation of manual therapists keen to display their philosophies and techniques. In addition, long-standing and established practitioners have been able to completely review their con­ tributions as the result of continuing practice and research. The practical application and scientific basis of manual therapy marches on. Clinical problem-solving has become part of every ther­ apist's repertoire and this, linked to the need for rigorous quality assurance measures, has increased the need for research to support the use of manual therapy in a cost con­ scious world. The authors of the chapters have all produced out­ standing work, which allows this book to remain at the forefront of physiotherapy practice. No doubt, by the time the next edition is produced yet another group of aspiring manual therapists will be ready to share their professional expertise. The progress of manual therapy moves ever onward. In conclusion, it is to be hoped that this text will be use­ ful to undergraduates, to practising manual therapists and to the ever-increasing number of therapists completing higher degrees.

London and Cardiff, 1994

J. D. B. N. P.

Preface to the first ed ition

Churchill Livingstone's invitation to compile and edit a text on Modern Manual Therapy prompted my first concept of a rich and comprehensive totality. Constraints of the possible soon whittled down that version, yet the chapters are, I hope, a fair representation of what physiotherapists were thinking and doing in the mid-1980s, together with author­ itative accounts of some contexts of that work. I have enjoyed the privilege of being associated with the sixty authors, whose views I may not necessarily share of course. Together with excellent contributions from British col­ leagues, the manifest overseas presence reflects my abid­ ing links with those energetic and restless countries whose citizens have contributed much sound, realistic advancement. This is not an exhaustive text on technique, nor even a rep­ resentative vocabulary. Technique is not of prime importance, since technique springs most naturally from the fullest grasp of the nature of the musculo-skeletal problem. More arduous than learning the various ways to push this or tweak and pull that is the task of educating oneself in understanding the problem. This is infinitely worthwhile and rewarding, because this also teaches when not to handle the patient. Improvement of clinical competence is a demanding business. Ultimately, clinical effectiveness is directly related to the strength of the individual's desire to be clinically effective, and it is pointless beseeching deaf heaven, 'Will somebody please tell me what to think', since always there are those only too happy to do this. Workers who seek to improve their clinical efficacy need discrimination and lively ability to distinguish fact from fancy. We derive from each other, as the painter Sickert (1860-1942) has expressed it: ' . .. the language of paint, like any other language, is kneaded and shaped by all the com­ petent workmen labouring at any given moment; it is, with all its individual variations, a common language and not one of us would have been exactly what he is but for the influence and experience of all the other competent work­ men of the period.' Many recent advances in basic knowl­ edge, and alternative ways of thinking about old problems, have already made our yesterdays seem centuries ago, yet

we need to recognize sterile propaganda and plain adver­ tisement. Novelty is not progress. By its nature, manipulative medicine does not enjoy the same scientific basis as anatomy, physiology, molecular biology, pathology or pharmacology, for example. We can­ not take the bits apart to see what we are doing, or why we need to do it. Much of what we do is simply what has been proven on the clinical shop floor to be effective in getting our patients better - we do not always know precisely why. We continue to sound as though we know so much, when we know comparatively little. It might be a good thing to admit to this. We make much of clinical science, enthusiastically referring to this or that part of the massive mountain of literature which best serves our particular interest, yet Oliver Sacks (1982), who researched the effects of L-dopa on Parkinson's disease, puts the matter clearly: 'We rationalise, we dissimilate, we pretend; we pretend that modern medicine is a rational science, all facts, no nonsense and just what it seems. But we have only to rap its glossy veneer for it to split wide open and reveal to us its roots and foundations, the old dark heart of metaphysics, mysticism, magic and myth.' As astrology is to the science of astronomy, pure science tends to fall by the wayside as wishful thinking, therapeu­ tic likes, dislikes and old loyalties push to the fore. While it is ordinary common sense to work in the way in which one feels most comfortable, and most effective, we cannot thereby make a scientific virtue out of expediency. Professor Lewis Thomas, of the State University of New York at Stony Brook, recently mentions (in Late Night Thoughts 1984 OUP): 'Medicine, the newest and youngest of all the sciences, bobs along in the wake of biology, indeed not yet sure that it is all that much of science, but certain that if there is to be a scientific future for medicine it can come only from basic biomedical research.' Manual therapists may have a long road to travel before we talk an agreed common language, founded on scientific fact, but we can enjoy some solid progress towards that end and are now travelling with confidence. Halesworth, Suffolk, 1986 G. P. G.


Publications of the size and quality of the third edition of Grieve's Modern Manual Therapy: The Vertebral Column can­ not come to fruition without the work of many individuals. As Editors, we would like to thank most particularly the contributors to this text.They not only gave of their time to write the chapters, but the written material presented in this text reflects the contributors' lifelong work and dedica­ tion to enhancing the sciences and clinical practices of today's modern manual therapies. The contributors are to be congratulated on their outstanding work, their impres­ sive research and cutting edge applications to clinical prac­ tices. The text represents literally hundreds of years of experience and reveals the leadership of physiotherapists in the musculoskeletal field.

Thanks are also given to the publishers, Elsevier, and in particular to Mary Law, Barbara Muir, Dinah Thorn, Claire Wilson and Ailsa Laing whose untiring work and, at many times, patience has brought this third edition to print. Stephanie Pickering is also to be thanked for her attention to detail in copy-editing the manuscript. Any errors remaining are naturally those of the Editors. Finally, we would like to acknowledge the tolerance of our respective families and friends. We thank them for their patience and support during the preparation of this publication.

J. D. B. G. A. J




Introduction to modern manual therapy

SECTION CONTENTS 1. The future scope of manual therapy 3





The future scope of manual therapy
J. D. Boyling, G. A. Jull

Among the many developments over the past decades in the field of spinal pain, two which are having a major impact on clinical practice in the field of manual therapy are: defining spinal pain within a biopsychosocial model (Waddell 1992) and the calls for, and moves towards, the adoption of evidence-based practices (Sackett et aI1997). Placing spinal pain in the context of a biopsychosocial model has improved understanding of the multidimen­ sional nature of pain and disability and has underpinned shifts and expansions in management approaches. Practising within this model has had undoubted benefits for back and neck pain patients. Nevertheless there are still challenges ahead. Even with the adoption of this model, there does not appear to have been any lessening in the life­ time incidence of neck and back pain, neither is there evi­ dence that there has been any substantial success in preventing the transition from an initial acute episode of pain to a recurrent or chronic pain state. One of the historic problems in this field has been the dif­ ficulty in obtaining a definitive patho-anatomical diagnosis for the vast majority of patients with an episode of neck or back pain. Working within a patho-anatomical model, researchers and clinicians still have to contend with such diagnoses as non-specific back pain, idiopathic neck pain, or neck pain following a whiplash injury. This in itself is unsatisfactory, but as is well appreciated clinically, posses­ sion of a definitive diagnostic label such as a 'discal injury' may not be much more helpful in directing treatment. Under such a diagnosis many different clinical presenta­ tions are possible, which often require different manage­ ment approaches. Given this situation, there are shifts in the paradigm of research in the medical literature. The shift is towards try­ ing to better understand the processes in the pain, neuro­ muscular and psychological systems underlying patients' pain, disability and functional problems and their interac­ tion. Health practitioners such as physiotherapists are well positioned to contribute to this research, as this is their model of practice. Historically, from the patient interview and physical examination, the manual therapy clinician has



aimed to understand the patient as a person and how their spinal pain is affecting them personally and functionally, and to elucidate the nature of impairments in the articular, muscular and neural systems that are associated with the patient's problem. It is therefore pleasing to observe that the basic and applied clinical sciences of manual or musculo­ skeletal therapy have undergone rapid development in the past decade in this mechanistic model of research. As is evi­ dent in the third edition of this text, researchers from the disciplines of manual therapy are involved in the basic and applied clinical sciences to better define the processes in spinal pain and disability. The outcomes of this research are indicating that quite specific problems occur in the various systems and the changes can be variable in nature and degree. Such changes in the pain and neuro-motor systems, with their attendant psychological responses, appear to occur simultaneously and interdependently. The outcomes of such mechanistic basic sciences research have the poten­ tial to indicate the type of treatment that is likely to be most suitable to reverse a certain problem. W hat has become evi­ dent from this research, and well known to clinicians, is that back and neck pain are not homogenous conditions. Researchers in the applied clinical sciences are testing the effectiveness and efficacy of these research directed inter­ ventions, but the current challenge is to better understand the precise nature of the changes and, most importantly, to be able to identify and classify the disorders and recognize patients who are more likely to be responsive to certain treatment approaches. The World Health Organization (W HO) has provided a starting point with two publica­ tions. The first is the International Statistical Classification of Diseases and Related Health Problems (ICD-IO) (World Health Organization 2003a). Second, and of more interest, is the International Classification of Functioning, Disability and Health (ICF) (World Health Organization 2003b). Manual therapy practices have changed over the past decade in response to new knowledge, and they will con­ tinue to undergo change and refinement. The continuing coalescence of the science and clinical practices of manual therapy will further strengthen the approach embraced in evidence-based practice. Not surprising, given the verifica­ tion of multisystem involvement in neck and back pain, the evidence is pointing towards the greater efficacy of multi­ modal therapies, particularly inclusive of exercise in the management of neck and back pain. However, the evidence of efficacy is not as yet unequivocal for any conservative management method and this is placing tensions on all in the healthcare sector internationally. Many reasons can be offered for this current state but perhaps the more impor­ tant need is future directions which will assist the advance­ ment of clinical evidence to assist the community to obtain optimal health care for neck and back pain. Clinicians need to play a major role and the nature of their participation is illustrated in Figure 1.1:

Evidence-based practice

Clinical guidelines for patient categories

Practice-based evidence

Role of clinicians Figure 1.1 practices. The clinician's contribution to evidence-based


The evidence gained from clinicians treating patients is an important driver for research, and for the further devel­ opment and implementation of evidence-based practices.

further identification of physical and psychosocial processes in spinal pain patients recognition of recurring patterns of processes diagnostic groups responses to interventions - evidence with patient­ centred outcomes and outcomes of physical impairment and functioning, documentation of relationships responsiveness to treatment - identification of responders and non-responders data on patients' values, experiences and opinions of treatments.

This means that musculoskeletal physiotherapists, be they clinicians or researchers, need to look at outcomes. An out­ come is that which comes out of something - a visible or practical result, effect or product. There are a number of fundamental questions. What should be measured? How do I measure the outcome? How do I use the measurement to analyse the efficacy or efficiency of the rehabilitation? The ICF provides a conceptual framework to understand the consequences of disease including spinal pain. The con­ sequences act at the level of impairment, activity limitation and participation as well as at the level of quality of life. Haigh et al (2001) have reported that the majority of out­ come measurement is at the impairment level, with some at the activity limitation level and very little at the quality of life level. It is worth remembering that musculoskeletal physiotherapy acts at more than the impairment level and therefore measures of outcome should reflect this. However, evidence-based practice is shaped by what forms of knowledge are counted as evidence. In view of this, Gibson & Martin (2003) have highlighted the role of


tative research in evidence-based physiotherapy practice.

The future scope of manual therapy


The destiny of manual therapy must be controlled by its clinicians, researchers and consumers. The third edition of

thought and appraisal to drive future research and clinical practice in manual therapy.

Modern Manual Therapy has changed direction from previ­
ous editions, to highlight developments in the field. It embraces the biopsychosocial model of back pain and evi­ dence-based practices and highlights the basic and applied clinical sciences underpinning current practices. Foremost, it should improve practice and open avenues for critical

KEYWORDS biopsychosocial model evidence-based practices practice-based evidence classification outcome

Gibson B E, Martin 0 K 2003 Qualitative research and evidence-based physiotherapy practice. Physiotherapy 89(6): 350-358 Haigh R, Tennant A, Biering-Sorensen F et al 2001 The use of outcome measures in physical medicine and rehabilitation in Europe. Journal of Rehabilitation Medicine 33: 273-278

Waddell G 1992 Biopsychosocial analysis of low back pain. Bailliere's Clinical Rheumatology 6: 523-557 World Health Organization 2003a International statistical classtfication of diseases and related health problems (lCO-10), 10th edn. World Health Organization, Geneva World Health Organization 2003b International classification of functioning, disability and health (ICF). World Health Organization, Geneva

Sackett 0 L, Richardson W S, Rosenberg W, Haynes R B 1997 Evidence­ based medicine: how to practice and teach EBM, 1st edn. Churchill Livingston, New York




Foundation sciences for manual therapy

SECTION CONTENTS 2. Comparative anatomy of the spinal disc 9 17

3. Comparative anatomy of the zygapophysial joints 4. Kinematics of the spine 31

5 . Chemistry o f the intervertebral disc i n relation t o functional requirements 6. Clinical biomechanics of the thoracic spine including the ribcage 7. Clinical biomechanics of the lumbar spine 8. Clinical biomechanics of lifting 89 105 67 55


9. Motor control of the cervical spine 10. Motor control of the trunk 119

11. The lumbar fasciae and spinal stability

141 155

12. Neurophysiology of pain and pain modulation 13. The effect of pain on motor control 14. The spine and the effect of ageing 173 187





Comparative anatomy of the spinal disc
s. Mercer

CHAPTER CONTENTS The intervertebral disc
9 10 10 12

Lumbar intervertebral disc Cervical intervertebral disc Thoracic intervertebral disc Blood supply Innervation
13 14 14

Clinical implications

The vertebral column acts as the central flexible rod of the trunk. Therefore each intervertebral disc, interposed between adjacent vertebrae, has several functions. Primarily it acts to separate the vertebral bodies allowing them to move relative to each other and thereby promoting motion at each interbody joint. Additionally, a disc must sustain the load of the body above it and the action of any surrounding muscles when they act. In order to carry out these functions an intervertebral disc must be pliable yet strong (Bogduk 1994). Each section of the vertebral column must also meet spe­ cific regional demands. The cervical spine must ensure bal­ ance and free movement of the head. The thoracic spine provides for suspension of the ribs and therefore support of the thoracic cavity. The lumbar spine, opposite the abdom­ inal cavity, ensures mobility between the thoracic portion of the trunk and the pelvis while withstanding the higher loads of the trunk above. The morphology of the vertebrae of each section of the spine reflects these regional differences in function. In the lumbar spine the superior and inferior surfaces of the ver­ tebral bodies are comparatively large and flat reflecting their load transfer function (Bogduk 1997). On the other hand, the superior surfaces of cervical vertebrae 2 -7 have uncinate processes reflecting the need for multidirectional mobility of the neck and also the need for stability (Penning 1988). The vertebral bodies of thoracic vertebrae 2 - 10 increase in size and change shape down the vertebral col­ umn and, importantly, each has two demi-facets for the attachment of ribs (Breathnach 1965). This association of the thoracic vertebrae with the ribcage results in a more rigid region of the spine (Takeuchi et a11999). The notion of form and function when considering the bony morphology of regional or individual vertebrae is not unusual as musculoskeletal physiotherapists are familiar with these changing shapes and sizes of vertebrae reflecting the regional changes in function within the vertebral column . Yet, when the morphology of the intervertebral



disc is considered, a fairly uniform structure is typically portrayed. The archetypal intervertebral disc is depicted as a nucleus pulposus encircled by an annulus fibrosus, inter­ posed between a superior and an inferior end-plate (Williams et al 1995). However, this description is based on the anatomy of a lumbar disc, the region where most research concerned with the spine has occurred and from which many authors have extrapolated the anatomy to all intervertebral discs. More recently, studies of the cervi­ cal intervertebral discs have demonstrated that these discs are distinctly different to lumbar discs and that these dif­ ferences are evident from birth (Mercer & Bogduk 1999, Oda et al 1988, Pooni et al 1986, Scott et al 1994, Taylor 1974, Tondury 1972 ). Little is currently available in the literature regarding thoracic disc morphology

In the lumbar region the nucleus pulposus consists of a cen­ tral core of proteoglycan matrix surrounded by fibrocarti­ lage. In infancy the nucleus pulposus is a soft gel and occupies three-quarters of the anterior-posterior dimension of the disc (Taylor et al 2 000). Although dehydrating with age, the healthy adult nucleus pulposus is still a semi-fluid mass of mucoid material. Taylor et al ( 2 000) found that even in cadaveric material of older adults the nucleus still demonstrates the ability to imbibe water (Fig. 2 .1). The lumbar annulus fibrosus consists of approximately 10-2 0 concentric lamellae of collagen fibres which surround the nucleus pulposus. Collagen fibres within each lamella run in parallel at an angle of approximately 65 degrees to the vertical but for each pair of lamellae the direction of the fibres alternates. Such an arrangement enhances the capac­ ity of the lumbar annulus to restrain different movements in diverse directions (Bogduk 1997). Alternating the direc­ tion of fibres in each lamella is vital in the disc resisting twisting (Hickey & Hukins 1980).

Typically the lamellae are depicted in diagrammatic form with each one completely encircling the nucleus pul­ posus and being of fairly uniform thickness. However, the thickness of each lamella varies with location and each one does not necessarily form a complete ring around the disc (Marchand & Ahmed 1990). The lamellae closer to the nucleus pulposus are thicker. Furthermore the anterior and lateral lamellae are thick while the posterior lamellae are thinner and more closely packed (Marchand & Ahmed 1990). When viewed from above the posterior portion of the lumbar annulus fibrosus is therefore narrower than the anterolateral aspects ( see Fig. 2 .1). Incomplete lamellae, that is lamellae that fail to pass around the circumference of the disc, are normal anatomy. They have been noted to be more common in the mid-portion of the disc (Tsuji et al 1993). Marchand & Ahmed (1990) report that within any quadrant of the disc about 40% of the lamellae are incomplete while in the posterolateral comers some 50% are incomplete. When incomplete, the lamella will fuse or approximate with the lamellae superficial or deep to it. On the basis of attachment sites two portions of the annulus fibrosus may be identified. The outermost lamellae insert into the ring apophysis of the upper and lower verte­ brae. These fibres, attaching bone to bone, may be consid­ ered as ligaments and as such are designed primarily to limit motion between adjacent vertebrae. The inner lamel­ lae do not attach to bone, rather they attach to the !?uperior and inferior cartilaginous end-plates. These more cartilagi­ nous, proteoglycan-rich lamellae form an envelope around the nucleus pulposus (Taylor et al 2 000) and so resist any radial expansion of it (Bogduk 1997). The cartilaginous end-plates bind the disc to the verte­ bral bodies and act in the transmission of load. They cover the central area of the vertebral body encircled by the ring apophysis. Closer to its vertebral surface the end-plate is composed of hyaline cartilage while its discal surface is fibrocartilage (Peacock 1951). The inner fibres of the annu­ lus fibrosus are strongly attached to the vertebral end­ plates while the end-plates are only weakly attached to the vertebral body. Consequently the end-plates are considered part of the intervertebral disc rather than as part of the lum­ bar vertebral body (Coventry 1969, Taylor 1975). Such mor­ phology renders the disc susceptible to avulsion from the vertebral body in some forms of trauma.

Figure 2.1

Photograph showing a top view of a 39-year-old

lumbar intervertebral disc. The annulus fibrosus surrounds the nucleus pulposus (NP).


is thick and

Detailed study of the normal cervical intervertebral disc has only recently been undertaken and the results indicate that the anatomy of the cervical disc is distinctly different to that of the lumbar intervertebral disc (Mercer & Jull 1996). From birth the nucleus pulposus of the cervical disc com­ prises a much smaller portion of the disc, some 2 5% rather than the 50% seen for the lumbar nucleus (Taylor 1974). In addition the nucleus, even in infancy and childhood, has a higher collagen content than the thoracic or lumbar nucleus -

Comparative anatomy of the spinal disc


(Scott et al 1994, Taylor et al 1992 ). Furthermore, by adoles­ cence or adulthood the nucleus is no longer mucoid in naturebut is characterized by fibrocartilage (Oda et al 1988, Tondury 1959, 1972 ). Bland & Boushey (1990) state that, by 40 years of age, there is no gelatinous nucleus pulposus; rather this central region of the cervical disc is composed of fibrocartilage, islands of hyaline cartilage and tendon-like material. Anatomical studies to date indicate that a gelati­ nous nucleus pulposus is only to be expected in children and young adults. The adult cervical nucleus pulposus is characterized by fibrocartilage (Fig. 2 . 2 ). Examination of the three-dimensional anatomy of the cervical intervertebral disc reveals that it does not mirror the morphology of the lumbar disc (Mercer & Bogduk 1999). The annulus fibrosus is not a ring-like structure of lamellae. Rather it is a discontinuous structure which com­ prises two distinct portions. The anterior annulus, found running anteriorly between the uncinate processes, is cres­ centic in shape. It is well developed and thick at the mid­ line, tapering laterally and posteriorly as it approaches the anterior margin of the uncinate processes (Fig. 2 . 2 ). The ori­ entation of the collagen fibres within the anterior annulus is also dissimilar to the lumbar annulus fibrosus. In the cervi­ cal disc the fibres of the anterior annulus converge superi­ orly towards the lower anterior edge of the vertebral body above. The anterior annulus may therefore be considered as an interosseous ligament, arranged like an inverted 'V' whose apex is located at the axis of axial rotation (Bogduk & Mercer 2 000, Mercer & Bogduk 2 001). What we may con­ sider the posterior annulus is a small structure represented by a few vertically oriented fibres located close to the median plane at the posterior aspect of the disc. It is a thin lamina, being no more than 1 rnrn in depth ( see Fig. 2 . 2 ). The posterolateral aspects of the cervical disc therefore lack

Figure 2.2

Photograph showing the top view of a 39-year-old

cervical intervertebral disc. The anterior annulus fibrosus (AF) is thick and fibrous, tapering posteriorly towards the uncinate region. Posteriorly the thin annulus fibrosus (AF) is found only towards the midline. Centrally the nucleus pulposus (NP) appears as a fibrocarti­ laginous core.

the support of an annulus fibrosus. Only the posterior lon­ gitudinal ligament covers the majority of the posterior disc. Posterolaterally the uncovertebral clefts are overlaid by periosteofascial tissue (Fig. 2 . 3). This unorganized fibrous connective tissue embedded with fat and a large number of blood vessels i's continuous with the periosteum of the ver­ tebral body and pedicles (Mercer & Bogduk 1999). Centrally, the nucleus pulposus of the adult cervical disc is fibrocartilaginous in nature (Bland & Boushey 1990, Oda et al 1988, Tondury 1972 ). The clefts, which extend into this fibrocartilaginous core, open under the periosteofascial tis­ sue (Mercer & Bogduk 1999). These clefts begin developing

Figure 2.3

Photograph of cervical intervertebral disc from behind. On the left the uncovertebral cleft'(UC) which extends into the fibro­

cartilaginous core. On the right the periosteofascial tissue (PF) which covers the uncovertebral cleft.



Figure 2.4

Photograph of a sagittal section through cervical

C2/C3 and C3/C4. Note the anterior annulus (AF) and narrower posterior annulus fibrosus (pAF) . The uncovertebral clefts (UC) have transected the posterior two-thirds
of the intervertebral discs.

intervertebral discs

Figure 2.5

Photograph of a coronal section through cervical

intervertebral discs. The section through the disc reveals the uncinate processes

C5/C6 intervertebral (UP) and uncovertebral cleft

(UC). The coronal sections through the higher discs are further

between 9 and 14 years of age when the uncinate processes reach their maximum height ( Ecldin 1960, Tondury 1959). With increasing age the clefts.penetrate more medially into the core until they completely transect the posterior two­ thirds of the disc, occasionally leaving a small isolated bar of fibrocartilage just deep to the posterior annulus ( Ecklin 1960, Mercer & Bogduk 1999, Tondury 1972 ) ( Figs 2 . 4, 2 . 5). These clefts are normal anatomy of a cervical disc which, together with the absence of a substantial posterior annu­ lus, facilitate axial rotation (Bogduk & Mercer 2 000, Mercer & Bogduk 2 001).

posterior and reveal the penetration of the clefts towards the midline to transect the posterior disc.

Very little is known of the detailed morphology of the tho­ racic intervertebral disc. Pooni et al (1986) reported that in cross-section thoracic discs were more circular than either cervical or lumbar discs, which were more elliptical in shape. In addition thoracic discs were less wedge-shaped. Although depicted in a variety of texts as similar in gross structure to lumbar discs ( Kapandji 1974, Woodbume & Burkel 1988), Zaki (1973) described the annulus fibrosus of the thoracic disc to be a discontinuous two-part structure, with the fibres of the posterior annulus being of vertical ori­ entation. He gave no indications regarding the transition of morphology from cervical to thoracic disc or thoracic to

lumbar disc, or of transitions within the thoracic spine. In addition Lee (1994) postulated the presence of transverse fissures in the thoracic disc. Recent preliminary work regarding the three-dimensional anatomy of the thoracic intervertebral disc has indicated that the thoracic discs through to the T9/TlO level exhibit a morphology similar to the cervical disc (Mercer 2 001). The anterior annulus fibrosus is crescentic, thicker anteriorly towards the midline, and tapering laterally and posteriorly to the costal region ( Fig. 2 . 6). The central fibres of the radi­ ate ligament pass horizontally anterior to the annUlus fibro­ sus, to be covered by the fibres of the anterior longitudinal ligament. Posteriorly the fibres of the thin, centrally placed posterior annulus fibrosus are vertical, being covered by the central longitudinal fibres and lateral extensions of the pos­ terior longitudinal ligament. Posterolaterally, fromTl/T2 to T9/TlO the head of the rib articulates with the upper and lower demi-facets and with the intervertebral disc via the intra-articular ligament ( Fig. 2 . 7). At these levels the anterior annulus has tapered prior to the costovertebral joints. Within the fibrocartilaginous core, fissures and clefts are ubiquitous and normal ( Figs 2 . 8, 2 . 9).

Comparative anatomy of the spinal disc


Figure 2.6

Photograph of a top view through a transverse section

of a T2/T3 intervertebral disc. The anterior annulus fibrosus (AF) is much thicker than the posterior annulus fibrosus (pAF) tapering lat­ erally towards the costovertebral joint (NP) is located centrally.


The nucleus pulposus

At lower levels, where the head of the rib is articulating with only one vertebral body and not with the disc, the tho­ racic intervertebral disc adopts a lumbar-type three­ dimensional morphology (Mercer 2 001). Beginning at the TIO /THlevel, the annulus fibrosus is free to pass around the circumference of the disc as seen in the lumbar spine ( Fig. 2 .10). Here the nucleus pulposus, upon sectioning, would show signs of swelling or weeping as has been reported for lumbar discs. The typical thoracic disc appears to have been adapted from a cervical design rather than from a lumbar design. The annulus fibrosus of the cervical intervertebral disc morphol­ ogy has a posterolateral deficiency where the rib can gain access to the fibrocartilaginous core without having to nego­ tiate a posterolateral annulus fibrosus. The transition occurs from this morphology to a lumbar disc morphology at the

Figure 2.7

Photograph of a top view through a transverse section

of a T5/T6 intervertebral disc. The anterior annulus fibrosus (AF) tapers as it approaches the costovertebral joint


to surround

the nucleus pulposus (NP) anteriorly and laterally.

level where � rib is no longer associated with the interver­ tebral disc and articulates solely with the vertebral body.

As there are no major arterial branches directly supplying each intervertebral disc, a disc may be considered as an

Figure 2.8

Photograph of an upper thoracic intervertebral disc from behind. On the left the periosteofascial tissue has been resected to

reveal the uncovertebral cleft opening beneath it.


On the right the periosteofascial tissue has been left in situ to demonstrate the uncovertebral cleft




underlying the end-plates and in the base of the vertebral end-plate, the terminal branches of the metaphyseal arter­ ies and the nutrient arteries of the vertebra form a dense capillary network. Nutrients are then able to diffuse through the permeable central portions of the vertebral end-plates ( Urban et a11978). In the cervical spine Oda et al (1988) observed calcifica­ tion within the cartilaginous end-plate to begin in early adulthood. These authors postulated that such a process leads to a reduction of the nutritional route through the ver­ tebral end-plates leading to the early fibrotic changes observed in the nucleus pulposus.

Figure 2.9

Photograph of a sagittal section through the upper

thoracic spine. Uncovertebral clefts (UC) are present posteriorly. The posterior (pAF) is very thin while the anterior annulus fibrosus (AF) is relatively thick.

avascular mass of cartilage nourished by diffusion from blood vessels around its perimeter ( Taylor et al 2 000). Nutrients must therefore diffuse through the annulus fibro­ sus or through the vertebral end-plate to reach the nucleus pulposus. As demonstrated in the lumbar spine, the outermost fibres of the annulus fibrosus receive small branches from the metaphyseal arteries, which are anastomosing over its surface (Maroudas et al 1975). In the subchondral bone

Extensive plexuses cover the anterior, lateral and poste­ rior aspects of all intervertebral discs. These plexuses arise from the sympathetic trunks, gray rami communi­ cantes, vertebral nerve and ventral rami and send nerve fibres which penetrate the outer annulus fibrosus at all levels of the spine ( Bogduk et al 1981, 1988, Groen et al 1990). Nerve fibres and nerve endings have been identified in the outer third to half of the lumbar annulus fibrosus (Ashton et a11994, Bogduk et a11981, Hirsch & Schajowicz 1952 , Malinsky 1959, Palmgren et a11999, Rabischong et al 1978, Roofe 1940, Taylor & Twomey 1979, Yoshizawa et al 1980). Much less work has been carried out elsewhere in the spine. In the cervical region, nerve fibres have been demon­ strated in the outer third of the annulus fibrosus ( Bogduk et al 1988) or less specifically in the outer layers (Ferlic 1963). A more extensive pattern of innervation was described by Mendel et al (1992 ) who reported the presence of nerve fibres throughout the annulus, particularly in the middle third of the disc. These three studies indicate that the cervi­ cal intervertebral disc, like the lumbar disc, is innervated. However, precise anatomy of this innervation is lacking. Based on these findings for the cervical and lumbar intervertebral discs and the presence of extensive plexuses covering all intervertebral discs ( Bogduk et al 1981, 1988, Groen et a11990), it is reasonable to assume that the thoracic intervertebral disc has a similar pattern of innervation. However, the precise anatomy of this innervation awaits further study. Current evidence for innervation of the tho­ racic discs lies in clinical studies where pain is evoked with provocation discography (Wood et a11999).

Figure 2.10

Photograph of a top view through a transverse

section of a T11 /T12 intervertebral disc. The annulus fibrosus (AF) is now surrounding the nucleus pulposus (NP). Note that the anterior section of the annulus fibrosus is thicker than the posterior section of the annulus fibrosus.

An appreciation of the differing anatomy of the interverte­ bral discs throughout the spine is important when develop­ ing clinical models. The models developed for the lumbar ' intervertebral disc, such as internal disc disruption, radial and circumferential annular tears and disc herniation ( Bogduk 1991, Moneta et al 1994, Vanharanta et al 1987), are based on the structure of the lumbar intervertebral disc. As

Comparative anatomy of the spinal disc


the structure and function of the cervical and thoracic intervertebral discs are different to the lumbar disc the models developed for injury or the mechanism by which pain is produced in the lumbar disc are therefore not neces­ sarily applicable to models developed for the cervical and thoracic discs.
Ashton I K, Roberts S, Jaffray D C, Polak S M, Eisenstein S M 1994 Neuropeptides in the human intervertebral disc. Journal of OrthopaedicResearch 12: 186-192 Bland J, Boushey DR 1990 Anatomy and physiology of the cervical spine. Seminars in Arthritis andRheumatism 20: 1-20 Bogduk N 1991 The lumbar disc and low back pain. Neurosurgery Clinics of North America 2: 791-806 Bogduk N 1994 Anatomy of the spine. In: Klippel J H, Dieppe P A (eds) Rheumatology. Mosby, Sydney Bogduk N 1997 Clinical anatomy of the lumbar spine and sacrum, 3rd edn. Churchill Livingstone, Edinburgh Bogduk N, Mercer SR 2000 Biomechanics of the cervical spine. I: Normal kinematics. Clinical Biomechanics 15: 633-648 Bogduk N, Tynan W, Wilson AS 1981 The nerve supply to the human lumbar intervertebral discs. Journal of Anatomy 132: 39-56 Bogduk N, Windsor M, Inglis A 1988 The innervation of the cervical intervertebral discs. Spine 13: 2-8 Breathnach AS 1965 F razer's Anatomy of the human skeleton. J&A Churchill Ltd, London Coventry M B 1969 Anatomy of the intervertebral disk. Clinical Orthopaedics andRelatedResearch 67: 9-15 Ecklin U 1960 Die altersveranderungen der halswirbelsaule. Springer Verlag, Berlin F erlic D C 1963 The nerve supply of the cervical intervertebral disc in man. Bulletin of the Johns Hopkins Hospital 113: 347-351 Groen G J, Baljet B, Drukker J 1990 Nerves and nerve plexuses of the human vertebral column. American Journal of Anatomy 188: 282-296 Hickey D S, Hukins D W L 1980Relation between the structure of the anulus fibrosus and the function and failure of the intervertebral disc. Spine 5: 100- 116 Hirsch C, Schajowicz F 1952 Studies on structural changes in the lumbar annulus fibrosus. Acta OrthopaedicaScandinavica 22: 184--189 Hirsch e, Ingelmark B E, Miller M 1963 The anatomical basis for low back pain. Acta Orthopaedica Scandinavica 33: 1-17 Kapandji I A 1974 The physiology of the joints. Vol 3: The trunk and the vertebral column. Churchill Livingstone, Edinburgh Lee D 1994 Manual therapy for the thorax: a biomechanical approach. DOPe, Vancouver Malinsky J 1959 The ontogenetic development of nerve terminations in the intervertebral discs of man. Acta Anatomica 38: 96-113 Marchand F, Ahmed AM 1990 Investigation of the laminate structure of lumbar disc anulus fibrosus. Spine 15: 402-410 Maroudas A, Nachemson A, StockwellR, Urban J 1975 Some factors involved in the nutrition of the intervertebral disc. Journal of Anatomy 120: 113-130 Mendel T, Wink C S, Zimny M L 1992 Neural elements in human cervical intervertebral discs. Spine 17: 132-135 Mercer SR 2001 Transitions between cervical and lumbar intervertebral disc morphology. In: Proceedings of the 12th Biennial Conference, Musculoskeletal PhYSiotherapy Australia 31, Adelaide Mercer SR, Bogduk N 1999 The ligaments and anulus fibrosus of human adult cervical intervertebral discs. Spine 24: 619-628 Mercer SR, Bogduk N 2001 The joints of the cervical vertebral column. Journal of Orthopaedic and Sports Physical Therapy 31: 174-182 Mercer SR, Jull G A 1996 Morphology of the cervical intervertebral disc: implications for manual therapy. Manual Therapy 1(2): 76-81


lumbar intervertebral disc cervica I intervertebra I disc thoracic intervertebral disc

annulus fibrosus nucleus pulposus

Moneta G B, Videman T, Kaivanto K et al 1994 Reported pain during lumbar discography as a function of annular ruptures and disc degeneration: a re-analysis of 833 discograms. Spine 19: 1968-1974 Oda J, Tanaka H, Tsuzuki N 1988 Intervertebral disc changes with aging of human cervical vertebra: from neonate to the eighties. Spine 13: 1205-1211 Palmgren T, Gronblad M, Virri J, Kaapa E, Karaharju E 1999 An irrununohistochemical study of nerve structures in the anulus fibrosus of human normal lumbar intervertebral discs. Spine 24: Peacock A 1951 Observations on the pre-natal development of the intervertebral disc in man. Journal of Anatomy 85: 260-274 Penning L 1988 Differences in anatomy, motion, development and aging of the upper and lower cervical disk segments. Clinical Biomechanics 3: 37-47 Pooni J S, Hukins D W L, Harris P F, HiltonR e, Davis K E 1986 Comparison of the structure of human intervertebral discs in the cervical, thoracic, and lumbar regions of the spine. Surgical Radiological Anatomy 8: 175-182 Rabischong P, LouisR, VignilUd J, Massare C 1978 The intervertebral disc. Anatomica Clinica 1: 55-64 Roofe P G 1940 Innervation of anulus fibrosus and posterior longitudinal ligament. Archives Neurology and Psychiatry 44: 100-103 Scott J, Bosworth T, Cribb A, Taylor J 1994 The chemical morphology of age related changes in human intervertebral disc glycosarninoglycans from cervical, thoracic and lumbar nucleus pulposus and anulus fibrosus. Journal of Anatomy 180: 137-141 Takeuchi T, Aburni K, Shono Y, Oda I, Kaneda K 1999 Biomechanical role of the intervertebral disc and costovertebral jOint in stability of the thoracic spine: a canine model study. Spine 21: 1423-1429 Taylor JR 1974 Growth and development of the human intervertebral disc. PhD T hesis, University of Edinburgh Taylor JR 1975 Growth of the human intervertebral discs and vertebral bodies. Journal of Anatomy 120: 49-68 Taylor JR, Twomey L T 1979 Innervation of lumbar intervertebral discs. Medical Journal of Australia 2: 701-702 Taylor JR, Scott J E, Cribb A M, Bosworth TR 1992 Human intervertebral disc acid glycosaminoglycans. Journal of Anatomy 180: 137-141 Taylor J, Twomey L, Levander B 2000 Contrasts between cervical and lumbar motion segments. CriticalReviews in PhYSical and RehabilitationMedicine. 12: 345-371 Tondury G 1959 La colonne cervicale, son developpement et ses modifications durant la vie. Acta Orthopaedica Belgica 25: 6 02-625 Tondury G 1972 The behaviour of the cervical discs during life. In: Hirsch e, Zotterman Y (eds) Cervical pain. Pergamon Press, Oxford Tsuji H, Hirano N, Ohsrurna H, Ishihara H, Terahata N, Motoe T 1993 Structural variation of the anterior and posterior anulus fibrosus in the development of human lumbar intervertebral disc: a risk factor for intervertebral disc rupture. Spine 18: 204-210 Urban J P G, Holm S, Maroudas A 1978 Diffusion of small solutes into the intervertebral disc. Biorheology 15: 203-223 Vanharanta H, Sachs B L, Spivey M A et al 1987 T he relationship of pain provocation to lumbar disc degeneration as seen by CT I discography. Spine 12: 295-298 2075-2079



Williams P L, Bannister L H, Berry M M et al 1995 Gray's Anatomy: the anatomical basis of medicine and surgery, 38th edn. Churchill Livingstone, Edinburgh Wood K B, SchelLhas K P, Garvey T A, Aeppli 0 1999 Thoracic discography in healthy individuals: a controlled prospective study of magnetic resonance imaging and discography in asymptomatic and symptomatic individuals. Spine 24: 1548-1555 WoodburneR T, Burkel W E 1988 Essentials of human anatomy. Oxford University Press, Oxford

Yoshizawa H, O'Brien J P, Thomas-Smith W, Trumper M 1980 The neuropathy of intervertebral discs removed for low-back pain. Journal of Pathology 132 : 95- 104 Zaki W 1973 Aspect morphologique et fonctionnel de l'anulus fibrosus du disque intervertebrale de la colonne dorsaIe. Archives Anatomie Pathologie 2 1: 401-403




Comparative anatomy of the zygapophysial joints
K. P. Si nger, J. J. W. Boyle, P. Fazey

17 18 19

Development of the zygapophysial joints Zygapophysial joint morphology Zygapophysial joint capsule response to injury Articular asymmetry
21 22 23 21

Normal zygapophysial joint function and

Zygapophysial joint mechanics Innervation pattern

Zygapophysial joint loading and injury Manual therapy considerations


The design specification for the human vertebral column is the provision of structural stability, affording full mobility, as well as protection of the spinal cord and axial neural tis­ sues. While achieving these seemingly disparate objectives for the axial skeleton, the spine also contributes to the func­ tional requirements of gait and to the maintenance of static weight-bearing postures. At a component level, the paired zygapophyses of the human vertebral column are synovial joints within the 'functional mobile segment'. This term was coined by the German radiologist Herbert Junghanns (Schmorl & Junghanns 1971) to represent the union of two adjacent vertebrae, their intervening intervertebral disc (IVO) and articulations formed between the posterior elements. The regulation of compressive, shear and tensile forces applied to this 'triad' of disc and paired zygapophysial joints defines its functional role within the skeletal sys­ tem, both at the segmental level and within the spine overall. Understanding the variable structure and function of the human zygapophysial joints is an important require­ ment in manual therapy during the assessment and man­ agement of individuals with mechanical spinal pain disorders. Although in life, function of the mobile segment cannot separate out consideration of the intervertebral disc, this chapter will focus primarily on the development, form, function and variations in zygapophysial joints throughout the vertebral column. In some literature, the zygapophysial joints are referred to as facets, interlaminar joints, or the grouped term, posterior elements, is used. The most cranial zygapophysial joints are located between the second and third cervical levels, and the most caudal at the level of the lumbosacral junction. For reference to the specialized anatomy of the suboccipital region as well as the atlanto-occipital and atlanto-axial joints, the compre­ hensive review by Prescher is recommended (Prescher 1997).




The ossification of the posterior arches occurs separately from the vertebral body centrum and disc (O'Rahilly et al 1980). The paired neural arches unite to enclose the spinal canal and cord, from which stem the respective superior articular processes (SAP) and inferior articular processes (lAP), plus mammillary processes (MP), transverse, and spinous processes (Reichmann 1971, Rickenbacher et al 1985). There is an organized appearance of primary ossifi­ cation centres for each vertebral element (Bagnall et al 1977), which proceeds in a caudal direction and is generally complete by the fourth month in utero (Christ & Wilting 1992). According to Med (1977), during gestation the artic­ ular surfaces of the thoracic zygapophysial joints are rela­ tively flat, with the cervical and lumbar joints showing greater rates of remodelling. Impairment in normal devel­ opment, often in the first 4 weeks of gestation, has been speculated to contribute to joint configuration anomalies (Med 1980), in addition to segmentation anomalies, which can result in hemivertebra and block vertebra (Christ & Wilting 1992, Saada et aI2000). The rudimentary zygapophysial joint cavity and capsule is complete in embryos of 70 mm crown-rump length, and by birth the lAP and SAP of the zygapophyes are incom­ pletely ossified (O'Rahilly et aI1980). During development the lAPs, projecting inferiorly from the inferolateral aspect of the neural arch, engage with their respective SAPs to provide a congruent, symmetrical coupling. In the lumbar spine, the SAP is typically J-shaped, producing a coronally orientated medial component which acts to resist anterior shear strain, and a longer, more sagittal posterior part which acts to constrain rotation or torsion applied to the segment (Adams & Dolan 1995). The posteromedial margin of the SAP is given by Reichmann to show the most marked change, in particular the formation of the sagittal joint expansion (Reichmann 1971). The ossification of the lateral margin of the SAP is protracted during the first year of life with the expanding lateral cartilaginous cap lost to ossification until the defini­ tive form of the SAP is achieved by 7-9 years of age. This lateral element comprises the MP and projects posteriorly from the SAP to offer attachment to the multifidus muscle, which then ascends obliquely and medially, via tendinous slips, towards the superior two vertebral spinous processes (Macintosh et aI1986). The secondary ossification centres, at the tips of each of the articular, spinous, transverse, mammillary and acces­ sory processes, variously fuse during the first two decades of life (Singer & Breidahl 1990), taking their direction and shape according to the tensile forces applied to them from the attaching musculature and ligaments (Lutz 1967). Indeed, anomalous development of the multifidus muscle, originating from the MP of the SAp, is given by Odgers (1933) to account for asymmetric configuration of lumbar zygapophyses - termed 'articular tropism' by some authors.

The early prenatal configuration of the spinal zygapophyses is essentially similar throughout the spine in that they are aligned predominantly on the frontal plane (Lewin et aI1962), although the precursors for their eventual adult form are already evident in some individuals (Reichmann 1971). During the first postnatal year, the shape of the paired zygapophysial joints changes as functional and regional demands are imposed. The specifications for the cervical and lumbar regions, through the relatively greater vertical dimension of the IVD, confer greater mobility on these segments. In contrast the thoracic discs, which account for only a fifth of the vertical dimension of this region, pre­ dispose less segmental sagittal plane motion (Gregersen & Lucas 1967). The regional variations in morphology of the cervical, thoracic and lumbar vertebrae and their respective zygapophysial joints are depicted in Figure 3.1. There is considerable variation in the alignment and shape of the zygapophyses throughout the spine, despite the tendency in modern anatomy textbooks to depict symmetry (Grieve 1981). At the transitional junctions, where developmental and pathological anomalies predominate (Schmorl & Junghanns 1971), there may be marked mor­ phological differences between right and left zygapophysial joints (Singer et a11989a) (Fig. 3.2). Even in areas remote from the transitional junctions there may also be marked joint asymmetry (Burkus 1988), providing an important caution against always inferring abnormal mechanical behaviour from passive motion assessment of spinal segments.

Figure 3.1 A series of a xia l, la tera l a nd posterior views of mid­ cervica l (A) , thora cic (B) a nd lumba r (C) vertebra e to depict the pri­ mary configura tion of their respective zyga pophysia l joints. In the cervical region, these joints lie la teral to the neura l a xis compared with the thora cic a nd lumba r joints. The typica l thora cic segment (B) shows the more vertica l a nd corona l a lignment wherea s the lumba r vertebra e (C) show the 'J'-shaped zyga pophyses with their corona l a nd sa gitta l elements.

Comparative anatomy of the zygapophysial joints


Four transverse CT images depicting articular asymme­ try, or tropism, of the paired zygapophysial joints. Where tropism occurs at one transitional junction, this and other anomalies may be found at adjacent transitions. The lower images are of a 35-year-old male, with a similar asymmetry pattern of Tl1-12 (C) and also at L4-5 (D).

Figure 3.2

SAP, there is typically a thicker cartilage in response to these lateral forces (Putz 1985) (Fig. 3.3).

Figure 3.1


The eventual adult configuration and shape of the zygapophyses is influenced by the exertional forces applied during early gestation and immediate postnatal motor development. Using in utero ultrasound, Boszczyk et al (2002) have speculated that prenatal morphological changes in zygapophysial joint shape occur in response to spinal torsion putatively induced from muscle actions. During early postnatal development, as the child adopts weight-bearing postures and commences crawling then walking, there is an intensified loading on the lateral mar­ gins of the joint which contributes to the sagittalization of the lumbar zygapophysial joints, as seen in the adult form (Lutz 1967). In the apex and lateral region of the lumbar

The articular surfaces are covered in hyaline cartilage and, like most synovial joints, have small fatty or fibrous syn­ ovial meniscoid-like fringes (Fig. 3.3) which project between the joint surfaces from the margins (Singer et al 1990). These intra-articular synovial folds (IASF) are found at all levels of the spine (Tondury 1972, Singer et al 1990, Mercer & Bogduk 1993) and are most developed within the polar regions, acting as space fillers during joint displace­ ments and actively assisting dispersal of synovial fluid within the joint cavity. Occasionally, the cartilage forms a non-articulating 'bumper' wrapping around the posteromedial aspect of the IAP of the joint, typically with a well-developed posterior expansion of the capsular ligament (Fig. 3.4). Often, these bumper cartilage formations are associated with evidence of articular cartilage degeneration and fissuring, ossification of the ligamentum flavum and reactive hyperplasia at the pos­ terior joint margins (refer to Fig. 3.5). The joint cavity is closed anteromedially and reinforced by the ligamentum flavum, which assists in approximation of the articular sur­ faces and, through its elastic properties, maintains the lumen



Figure 3.3 Photomicrograph of 100 11m thick transverse sections cut in the plane of the superior vertebral end plate at T11 - 12 showing a long, finger-like intra-articular synovial protrusion formed within the medial joint cavity, filling this void (A) . In the T12-L 1 joint (Bl. a fibro-fatty fold arising from the ligamentum flavum is depicted in the medial joint space projecting between the articular surfaces. In this instance, the SAP forms into an extended mammillary process, which wraps around the lAP. Note the uniform appearance of articular cartilage on all facets, with normal chondrocyte density evenly distributed, particularly with the apex of the lumbar joint (B). Adapted from Singer et al 1990. (C - articular cartilage; MP - mammillary processes; SAP - superior articular process; lAP - inferior articular process; LF - ligamentum flavum.)

of the vertebral canal (Ponseti 1995). Considerable ossifica­ tion within the ligamentum flavum may be associated with degeneration of the articular triad, although this tends to predominate in the region of the lower thoracic and upper lumbar segments (Malmivaara et al 1987, Maigne et al 1992). The articular processes of all zygapophysial joints com­ prise a cortical exterior containing trabecular bone with a thick subchondral region immediately adjacent to the artic­ ular cartilage. In regions of highest loading, for example the apex of the concavity of the biplanar lumbar SAP of the zygapophysial joints (see Fig. 3. 3), the subchondral bone is most dense, in response to shear and torsional loading. In contrast, the more planar joints of the cervical and thoracic regions tend to show a uniform distribution of cartilage across the face of the facet (Fig. 3.4). The articular cartilage is approximately 1 mm thick with a smooth surface in a normal articular facet. There may be regions of chondrocyte aggregation with thickening at zones of highest joint stress (see Fig. 3.3B). Reactive changes may be identified within the cartilage as a result of minor injury or degenerative changes. Complete enurbation of the cartilage is relatively rare given the tendency for repair via hyperplastic changes within the joint and its constituents which delay direct joint debridement (Fig. 3.5). In the cervical spine, the zygapophysial joints are rela­ tively flat while progressively increasing their surface area, and tend towards 45 degrees to the horizontal (see Fig. 3.1A), which reflects an increased axial loading of the head through the lower part of the cervical lordosis (Pal & Routal 1986). In the thoracic region, the joints adopt an almost ver­ tical direction while remaining essentially in a coronal ori­ entation (see Fig. 3.1B), which facilitates axial rotation and resists anterior displacement (Gregersen & Lucas 1967). The zygapophysial joints in the lumbar spine are vertical, with

Figure 3.4 Typical histological features of thoracic and lumbar zygapophysial joints where the ligamentum flavum encloses the joint space medially and the lateral joint margin is closed by the capsular ligaments. The relative differences in capsular ligament thickness is noted with the thoracic joint (A) depicting a slight, loose arrangement, which accommodates the excursion of the SAP on the lAP during rotation displacements (A). Both sections illus­ trate healthy articular surfaces despite slight incongruity of the lumbar joint, which also demonstrates a bumper extension of the articular cartilage wraps around the lateral margin of the lAP (B) . The respective elements labelled on the right. (AC - articular carti­ lage; MP - mammillary processes; SAP - superior articular process; lAP - inferior articular process; LF - ligamentum flavum; SB - subchondral bone; B - bumper cartilage; C - capsule.)

Comparative anatomy of the zygapophysial joints


Figure 3.5 Photomicrograph of a 100 11m-thick transverse section cut in the plane of the superior vertebral end plate at L1-2 to highlight unilateral zygapophysial joint degeneration. A normal intact joint is shown in the upper inset figure (A) and, in contrast, the higher magnification of the right joint (B) shows histological evidence of focal degeneration adjacent to a subchondral bone cyst and remodelling of the coronal region of the joint. Hyperplastic reactive bumper cartilage on the posterior margin of the lAP with thickening of the capsular ligament is also evident. (H - articular cartilage; lAP - inferior articular process; LF - ligamentum flavum; Be - bone cyst.)

tion in the thoracic region and composite motions in the cer­ vical spine. In a fresh, unpreserved lumbar spine, with the zygapophysial joints sectioned horizontally at the level of the superior vertebral end-plate, the ligamentum flavum and posterior joint capsular ligaments hold the articular sur­ faces firmly apposed. Where disc or zygapophysial joint injury or degeneration is apparent there is often greater joint play, unless the degenerative change is advanced. The liga­ mentum flavum is a substantial structure which envelops the anterior aspect of both the lAP and SAP (see Fig. 3.4), and maintains their approximation. The ligamentum flavum has two primary fibre orientations. Fibres are princi­ pally orientated vertically between adjoining laminae, although some pass medially and obliquely onto the ante­ rior aspect of the SAP, helping to form the posterior margin to the intervertebral foramen. Given the high proportion of elastin in this ligament (Tan et al 2003), its function is to maintain the lumen of the posterior wall of the vertebral canal and aid in elastic recoil of the spine back to its resting position, particularly after flexion motion (Ponseti 1995). The posterior joint capsule may merge its attachment into the peripheral articular boundary of the SAP, and in turn is reinforced by the tendinous slips of multifidus, which can tension the posterior joint. Occasionally, small sections of the posterior articular cartilage appear to become displaced from the subchondral bone (Taylor & Twomey 1986), possibly arising from sudden shearing of the lAP across the SAP under compressive or torsional load. Such examples of minor internal derangement of the zygapophysial joints respond well to manual therapy.

a curved, J-shaped surface predominantly in the sagittal plane (see Fig. 3.1C), which restricts rotation and also resists anterior shear. The change in shape of these joints between segments is generally progressive, although in some individuals there may be a more abrupt transition at the junctions between regions (Cihak 1981, Singer et al 1989a, Boyle et aI1996).

The morphology of the synovial joint capsule varies across the spinal regions. In the lumbar joints the capsule is thick and strong posteriorly to moderate sagittal plane move­ ments and resist torsion and extreme lateral flexion. This is in contrast to thoracic and cervical joints where it has a less robust composition (see Fig. 3.4) to permit the greater joint translations which occur in these regions, particularly rota-

Early descriptions of the role of the zygapophysial joints have defined their function as guides to direct and con­ strain segmental motion (Humphry 1858), a view endorsed by contemporary reviews of spinal biomechanics (Stokes 1988, Adams et al 2002). One of the more interesting per­ spectives on the functional role of the zygapophysial joints comes from the Canadian orthopaedist Harry Farfan, who conceptualized the 'spinal engine' (Farfan 1973). This mechanistic model employs the zygapophysial joints as cogs in a transmission to reciprocally transmit axial torque, generated by swinging the arms and shoulders, through the spinal segments to power the lower limbs for ambula­ tion (Farfan 1995). The cardinal role of the zygapophysial joints is to moder­ ate the direction and extent of segmental motion which may be safely sustained. As regional spinal motion capacity is regulated also by the shape and height of the intervertebral disc, an intrinsic role of the zygapophysial joints is protec­ tion, especially against excessive torsion and shear (Pearcy 1997). Shear strain is a major force vector in the lower lum­ bar segments given the lumbosacral angle, hence the poten­ tial for the initiation of spondylolysis, which can develop



through high compressive loading or repetitive dynamic loading (Sward et al 1991). Thus the zygapophysial joints can act both to facilitate and to limit physiological motion. Segmental axes of rotation vary correspondingly throughout the vertebral column moderated by the lor­ dotic or kyphotic alignment and the physical shape and height of the intervertebral discs. At the thoracolumbar junction (TLJ) the interlocking morphology of the zygapophysial joints (Singer 1989), coined a 'mortice joint' by Davis (1955) (Fig. 3. 6), limits motion mainly to sagittal plane movements and small gliding displacements. Caution is required by manual therapists when mobilizing TLJ and upper lumbar segments where rotation mobiliza­ tion and manipulation may be strongly countered by the 'mortice-type' configuration of the zygapophysial joints (Singer 1989, Singer & Giles 1990).

Articular asymmetry, or 'tropism', of spinal joint facets has been attributed in earlier reports to left or right hand dom­ inance (Whitney 1926), which may bias the movement pref­ erences and body directions in which an individual habitually moves. Others have suggested this may be caused through imbalance in muscle actions exerted against the joint (Odgers 1933, Lutz 1967). The incidence of

tropism of spinal joints is highest at the TLJ (see Fig. 3.2), typically the Tll-12 level, where 41% show>10 degrees of difference and 19% show >20 degrees of horizontal plane variation (Singer et al 1988) (Fig. 3.7). Similarly, at the cer­ vicothoracic junction (CTJ) almost a quarter of C6-7 joint pairs showed differences>10 degrees, while for C7-Tl and Tl-2 the differences were 18 and 16% respectively (Boyle et aI1996). In contrast, asymmetry is less common in the lum­ bar zygapophysial joints; however, at the lumbosacral junc­ tion articular tropism may be demonstrated. Cihak has reported up to 10 degrees of asymmetry in 16% of cases (Cihak 1970), and several other reports have confirmed this tendency (Putti 1927, Cihak 1981, Kenesi & Lesur 1985). Farfan has proposed that there was a higher incidence of unilateral lumbosacral NO prolapse on the side of the more coronal facing facet, which is disposed to torsion, compared to the side protected by a sagittal facing joint (Farfan et al 1972, Farfan 1983). In some individuals, tropism may have a developmental origin, whereas in others an acquired facet tropism may occur following injury to the zygapophysial joint resulting in remodelling. However, considerable variation in the ori­ entation and symmetry of the lumbar zygapophysial joints has been described in asymptomatic individuals, with much conjecture as to whether this contributes to late prob­ lems. As the lower lumbar motion segments are more fre-

Figure 3.6 Photomicrogra ph of a 100 11m thick tra nsverse section cut in the plane of the superior vertebra l end plate a t T11-12 illustra t­ ing a type I bila tera l mortice joint (A) formed by the embra cing ma mmilla ry processes which norma lly fuse with the la tera l expansion of the superior a rticula r process. Despite the a rticula r a symmetry, the hyaline ca rtila ge a ppea rs normal. A bila tera l mortice type joint configura tion a t T12-L1 is depicted with both ma mmilla ry processes forming a n enclosure to the respective lAPs (B) . Note the uniform a ppeara nce.of a rticula r ca rtilage on a ll fa cets. A fronta l plane CT image (C) demonstra tes the media l ta per effect of the lAPs, which would a chieve a com­ plete 'close-pa cked' position in a xia l weight-bea ring postures a nd extension of these upper lumba r zyga pophysia l joints. Ada pted from Singer 1989. (AC - a rticula r ca rtilage; MP - ma mmilla ry processes; SAP - superior a rticula r process; lAP - inferior a rticula r process; LF - liga mentum fla vum.)

Compa ra tive a na tomy of the zygapophysia l joints


140 120 100 80
.. .. '" " "



40 20

c;, c

r. Q. 0 Q. co

0 C :2. 140 "i Oii 120 ,.. 100 80 60 40 20 0 right zygapophysial joint angle

i:' c .!!

l5-S1 zygapophysial joint angles [Degrees)

Figure 3.7 The grea t va ria bility of thora columba r tra nsitiona l zyga pophysia l joint configura tions is clea rly evident in plots of the right vs left joint a ngles a t Tl1-12 a nd to lesser extent a t T12-L1. The ra nge of lumbosa cra l joint a ngles recorded by C iha k (1970) is depicted in the lower graph with the la rgest ra nge of joint a ngles a pproxima ting the corona l compa red to the sa gitta l plane. Ada pted from Singer et a l 1989a a nd C iha k 1970.

quently affected by injury and degeneration, joint tropism has been implicated as a possible aetiological feature. Cyron & Hutton (1980) observed that, when subjected to posteroanterior shear, motion segments with asymmetrical zygapophysial joints tend to rotate towards the more coro­ nally aligned joint. Manual therapy passive motion seg­ ment testing requires a· preparedness to accept that not all aberrant motion reflects underlying pathology (Grieve 1981). This reinforces the inadequacy of isolated testing and the necessity to consider all assessment findings, including imaging where available.

In the middle to lower cervical regions, the dual require­ ments of stability and mobility are provisioned through zygapophysial joints, which permit a composite of sagittal and lateral plane motions (Milne 1993b), with C5-6 con-

tributing the greatest segmental mobility (Fig. 3.8). The middle segments have a zygapophysial joint angle of approximately 45 degrees to the long axis of the spine, which reduces more abruptly at the CTJ (Boyle et alI996). The more caudal segments approaching the CTJ show a tendency for a smaller range of motion as the zygapophy­ ses adopt a form more characteristic of the upper thoracic segments. It is here that axial loading is higher and the seg­ mental mobility becomes markedly diminished as the tho­ racic cage commences (Bullough & Boachie-Adjei 1988, Boyle et al 1998). It is not unexpected that, with such an abrupt functional change at this transitional junction, severe fracture-dislocation injury can occur at this site, par­ ticularly in response to excessive applied forces as occur in motor vehicle roll-over accidents (Boyle et a12004). The uncinate processes, a unique feature of the cervical spine, whose form continues in the thoracic region as the paired costovertebral joints (Milne 1993a), strongly influ­ ences composite segmental motion, helping to prevent trans­ lation and, to some extent, lateral flexion (Bland & Boushey 1990, Milne 1991). The axes of rotation are commonly reported to be in the anterior region of the subjacent vertebra, with axial displacements progressively reducing towards the CTJ, corresponding with the change in inclination of the zygapophysial joints (Boyle et al 1996, 1998). In flexion, the upper cervical vertebra tilts and glides over the subjacent vertebra like an egg rolling in an egg cup. The composite cer­ vical spine motion is represented in Figure 3.8 both schemat­ ically, from multiple CT slice superimpositions, and graphically from ex vivo cadaver studies (Milne 1993a). The consequence of increased segmental mobility is the tendency for higher levels of disc degeneration (Singer 2000). Due to the oblique orientation of the cervical articular facets, the movements of rotation and lateral flexion are coupled within the cervical spine so that rotation is accom­ panied by ipsilateral lateral flexion. This motion can be con­ sidered to occur about a single axis, which is perpendicular to the plane of the zygapophysial joints as seen in the lat­ eral projection (Penning & Wilmink 1987, Milne 1993b). As the lower cervical and thoracic articular facets become more vertical, the axis of coupled motion could be expected to become more horizontal, involving more lateral flexion. However, the interfacet angles have been shown to have a bearing on the axis of coupled motion (Milne 1993b). At C3 and C4 the interfacet angles are less than 180 degrees and the orientation of the axis of coupled motion is constrained to a narrow band perpendicular to the facets (see Fig. 3.8); while in the lower cervical and thoracic regions, where the interfacet angles are greater than 180 degrees, the orienta­ tion of the axis of coupled motion can vary greatly depend­ ing on whether the applied force was axial rotation or lateral flexion. The articular surfaces of the cervical vertebrae not only regulate the direction and type of movement but, because of their oblique inclination, in a posteroanterior direction they also transmit the weight of the head (Med 1973). With



Figure 3.8 A reconstruction ba sed on functiona l CT studies, to show the na ture of composite rotation a nd side flexion occurring between the first cervica l a nd the first thora cic segments (A) . The a xes of coupled la tera l flexion a nd a xia l rota tion in the cervicothora cic spine (C2-T2 ) a re depicted schematica lly. Solid lines indica te the a xes of coupled motion when the a pplied force was rota ry, a nd the interrupted lines indicate the a xes when the a pplied force wa s la tera l bending. The lower three segment a xes shown can take on a wide ra nge of orien­ ta tion, but the range of motion here is quite limited in contra st to the middle three segments which have the widest potentia l excursion. Ada pted from Penning 8: Wilmink 1987 a nd Milne 1993b.

age-related changes in adult cervical spine posture, the load transfer role of the zygapophysial joints becomes increas­ ingly important in resisting anterior shear (Boyle et aI2002). The zygapophyses of the upper thoracic spine show some morphological features of the cervical region (Med 1972, 1973), and similarly the joints of the lower thoracic spine progressively approximate those of the upper lumbar region (Singer et aI1989a). The middle segments of the tho­ racic spine appear designed for less mobility as the thoracic cage articulations limit sagittal plane motion while accom­ modating axial displacements (Gregersen & Lucas 1967). The orientation of the articular facets in the thoracic spine changes only slightly throughout the middle region, approx­ imating the coronal plane and thereby permitting some sagittal motion and axial rotation while, in concert with the thoracic cage, limiting lateral bending. The middle thoracic segments, according to measurements involving pin inser­ tion into the spinous processes, showed the largest axial displacements compared with the upper and lower seg­ ments (Gregersen & Lucas 1967). There is an abrupt decrease in the range of axial rotation at the level of the TLl, as the zygapophysial joints conform to the typically sagittal configuration of the upper lumbar region (Malmivaara et al 1987, Singer et aI1989c). The axis of rotation for the thoracic spine has been described by Davis (1959) to lie in the region of the upper subjacent vertebral body, given the slight vertical inclina-

tion of the articular facets. In extension, the inferior pole of the articular facets can contact the laminae of the vertebra below which is believed to denote an important axial load transmission mechanism (Pal & Routal 1987). At the TLl, there is a specialized mortice-like arrangement, which appears in weight-bearing positions, designed to embrace the IAPs into the recess formed by the paired SAPs (Singer 1989) (see Fig. 3.6). This anatomical lock is accentuated by the medial taper of the SAPs into which the tenon-like IAPs fit (Fig. 3.6C). The zygapophyses of the lumbar spine are morphologi­ cally designed to prevent forward translation while allow­ ing considerable sagittal plane and lateral bending motions. The characteristic function of the lumbar spine is to trans­ mit axial load while providing stability and mobility of the trunk in relation to the lower limbs. A principal role of the upper lumbar zygapophysial joints is limitation of axial displacements (Fig. 3. 9), in part to protect the disc from tor­ sion (Farfan 1969), and to prevent anterior shear strain (Adams et aI2002). This requirement is well achieved in the upper lumbar spine, witnessed by the low rates of disc degeneration, prolapse or lis thesis, in contrast to the lower segments where disc injury is one consequence of the increased capacity for torsional displacements or increased shear in response to listhesis. Relative to disc height, there is a progressive increase in lumbar segmental mobility with the L4-5 and 1,5-S1 seg-

Comparative anatomy of the zygapophysial joints


Figure 3.10

Figure 3.9

A series of functiona l CT images a t L4-S to compa re the neutra l a nd subsequent side posture rota tion images of the same segment which highlights the ipsila tera l compression of the tension joint with sepa ra tion of the opposite side. The sca n pla ne wa s referenced to the superior vertebra l end plate a t L4 (A-D) . The typica l cha nge in configura tion of the lumba r zyga pophysia l joints describes the more sa gitta l orientation in the upper region, espe­ cia lly L1-2, to a progressively more corona l configura tion a t LS-S1.

From in vivo functiona l CT of the thoracic a nd lum­ ba r spine in norma l subjects; there were distinct differences evident a ccording to different zyga pophysia l joint morphologies, with evi­ dent a xia l displa cement of the TlO-11 thoracic segment (A) com­ pa red with the L4-S lumba r segment (B) . In contra st the upper lumba r segments with sa gitta l zyga pophyses show little differences from right or left rota tion postures (C ) , wherea s a t L4-S there is a greater tendency for ipsila tera l compression a nd sepa ra tion during side posture rota tion scans (D). Ada pted from Singer et a l 1989 a nd Singer et a l 2001.

ments contributing the most to sagittal plane motion. Through the tendency in the caudal segments towards more coronally angled . lumbar facets, slightly more axial plane motion may be achieved (Singer et al 200l) (Fig. 3.10). The anterior longitudinal ligament, which acts to passively constrain the lordotic postures, is a particularly well­ developed structure in lumbar and cervical regions, more so than its posterior counterpart. The classic work of Rolander (1966) demonstrated that the axes of rotation in the sagittal plane are principally located in the anterior region of the disc. For axial displacements, the axis of rota­ tion tends to be located within the posterior annulus. The

morphological adaptation of the last lumbar vertebra acts to allow torsion, by the more coronal orientation of the zygapophyses, as a requirement for locomotion (Boszczyk et al 2001). One consequence of segments disposed to excess torsion is the tendency for higher rates of disc degeneration (Farfan & Sullivan 1967, Farfan 1969, Singer 2000). In extension, the zygapophysial joints tend towards a close-packed position due to the apposition of the articular surfaces and the approximation of the inferior articular facet into the lamina below (Adams et al 1994). No differ­ ence was found in the range of lumbar rotation when sub­ jects were tested in full flexion, compared to upright standing, although the range of rotation increased when tested in a mid-position (Pearcy & Hindle 1991). The rota­ tional stiffness of an isolated motion segment is decreased by 40-60% following removal of the posterior elements (Markolf 1972). This emphasizes a key role of the lumbar zygapophysial joints in resisting rotation.




The physiological 'S'-shaped curve of the human spine con­ tributes to stability and to shock absorption, particularly during locomotion, in a manner analogous to a spring. However, the capacity for loading of these small joints varies depending upon their location. The cervical and lumbar zygapophyses are close to the line of gravity and consequently they contribute more to axial load transfer than the thoracic facets, which lie posterior to this line. This mechanical role of the zygapophyses and laminae as load­ bearing constructs has been examined as a function of sagit­ tal curve. Where the curvature is concave posteriorly, as in the cervical and lumbar regions, greater load was found to pass posteriorly (Pal & Routa11986, 1987). Ex vivo mechan­ ical studies of lumbar segments have confirmed that between 25 and 70% of the vertebral compressive load could be transmitted across the zygapophysial joints between adjacent vertebrae (Adams & Hutton 1980, Yang & King 1984). Sustained or dynamic compressive loading through the zygapophysial joints can increase significantly in loaded lordotic postures (Adams et al 2002), particularly those adopted in sports such as gymnastics and cricket bowling actions. In contrast, flexion loads are passed more anteriorly through the IVD, leaving the zygapophysial joints relatively unloaded. In this situation, anterior shear is resisted by the coronal portion of the SAP, which acts to prevent the forward displacement of the IAP. Such an anatomical restraint to flexion is important, as in full flexion there is quiescence of the extensor musculature (Kippers & Parker 1985). There are typical sites where function is disturbed when excess force is applied, as in the case of spinal injury result-

ing in fracture dislocation. Often, such injury is focused at locations of greatest morphological change between regions (Singer et a11989a, Boyle et al 2004), where the anatomy is least capable of dissipating the stress loading. The greater joint play associated with zygapophysial or disc injury has important implications for the concept of clinical instability. In the absence of reduced passive movement and symp­ toms consistent with instability, treatment decisions must be made with regard to the appropriate use of passive ver­ sus active stabilizing interventions.

The typical innervation pattern of the zygapophysial joints, lying so close to the spinal nerves, is via medial branches arising from the dorsal ramus, one of which descends around the SAP beneath the mammillo-accessory ligament to the inferior aspect of the same joint, with a descending branch to the superior aspect of the zygapophysis below (Groen & Stolker 2000) (Fig. 3.11). Thus each joint has a dual innervation, which is discretely unilateral in contrast to ventral structures, which possess a complex overlapping and bilateral innervation system (Groen & Stolker 2000). The zygapophysial joint capsule and IASFs (Giles & Harvey 1987) share this innervation, which may explain some types of segmental localized back pain syndromes which may be ameliorated by manipulation (Tondury 1971). Spasm of the multifidus muscle can be invoked with articular injury or entrapment of IASFs, given their shared innervation by branches of the dorsal ramus (Bogduk 1983, Bogduk & Marsland 1988, Groen et al 1990, Bogduk & Valencia 1994). The zygapophysial joints are therefore deter­ minants of both quality and quantity of lumbar spine move-

Figure 3.11 Horizonta l plane section of the mid-cervica l spine to illustra te the topographic a na tomy of the pa ired zyga pophysia l joints (Z) situa ted in the pla ne of the vertebra l ca na l. The spina l cord, dorsa l root ga nglia (*) a nd the emerging spina l nerves a re clea rly depicteo (A) . A schema tic illustra tion to depict the innerva tion of the pa ired zyga pophysia l joints from the media l (M) bra nches of the dorsa l ra mus. The intermedia te (I) branch supplying prima rily muscle a nd the la tera l (L) branch becoming cuta neous. Sympa thetic trunk (ST) . Permission to use these images wa s kindly provided by Professor Gerbra nd Groen, MD, PhD, Universitat Utrecht, a nd represent work in progress on the Huma n Spine CD project.

Comparative anatomy of the zygapophysial joints


ments and are an important source of local and referred low back pain (Mooney & Robertson 1976, McCall et aI1979).

The manual therapist commonly encounters zygapophysial joint related disorders in routine practice. As such, a clear understanding of their anatomy as it relates to clinical pres­ entation is necessary as an aid to forming a diagnosis and classification before evaluating the most appropriate course of action. For example, zygapophysial joint orientation may contribute information relevant to clinical presentation. The sagittal orientation of the posterior part of lumbar zygapophysial joints, along with the posterior capsule, restrains rotation to afford protection to the disc. Forceful rotation may therefore dispose the articular cartilage and subchondral bone to compression injury, particularly in the lordosed or extended position when the articular processes are more fully engaged. As well, the posterior capsule may be injured. Clinically, symptoms may then be reproduced by applied forces and combinations of movements that either compress the injured joint surfaces, for example extension and/or ipsilateral lateral flexion, or stretch the capsule via flexion and/or contralateral lateral flexion. Compressive patterns of pain reproduction may therefore
Adams M A, Dolan P 1995 Recent advances in lumbar spinal mechanics and their clinical significance. Clinical Biomechanics 10: 3-19 Adams M A, Hutton W C 1980 The effects of posture on the role of the apophyseal joints in resisting intervertebral compressive forces. Journal of Bone and Joint Surgery 62-B: 358-362 Adams M A, McNally D S, Chinn H, Dolan P 1994 Posture and the compressive strength of the lumbar spine. Clinical Biomechanics 9: 5-14 Adams M, Bogduk N, Burton A K, Dolan P 2002 Biomechanics of back pain. Churchill Livingstone, Edinburgh Bagnall K M, Harris P F, Jones P R M 1977 A radiographic study of the human fetal spine. 2. The sequence of development of ossification centres in the vertebral column. Journal of Anatomy 124: 791-798 Bland J H, Boushey D R 1990 Anatomy and physiology of the cervical spine. Seminars in Arthritis and Rheumatism 20: 1-20 Bogduk N 1983 The innervation of the lumbar spine. Spine 8: 286-293 Bogduk N, Marsland A 1988 The cervical zygapophysial joints as a source of neck pain. Spine 13: 61�17 Bogduk N, Valencia F 1994 Innervation and pain patterns of the thoracic spine. In: Grant R (ed) Physical therapy of the cervical and thoracic spine, 2nd edn. Churchill Livingstone, Edinburgh, pp 77-88 Boszczyk B M, Boszczyk A A, Putz R V 2001 Comparative and functional anatomy of the mammalian lumbar spine. Anatomical Record 264: 157-168 Boszczyk A A, Boszczyk B M, Putz R V 2002 Prenatal rotation of the lumbar spine and its relevance for the development of the zygapophyseal joints. Spine 27: 1094-1101 Boyle J J W, Singer K P, Milne N 1996 MorpholOgical survey of the cervicothoracic junctional region. Spine 21: 544-548 Boyle J W W, Milne N, Singer K P 1998 Clinical anatomy of the cervicothoracic junction. In: Giles L, Singer K (eds) Clinical anatomy and management of cervical spine pain. Butterworth Heinemann, Oxford, pp 40-52

be suggestive of zygapophysial joint articular cartilage involvement while stretch patterns may be more suggestive of capsular strain. This identification of the source of symp­ toms has implications for management with regard to encouragement of movement either towards or away from the pain-provoking direction. The same principles can be applied to cervical and thoracic regions with consideration of the movements constrained by either capsular tightness or articular process apposition. Effective manual therapy utilizes clinical application of knowledge of zygapophysial joint form and function. Formulation of a diagnosis based upon the clinical reason­ ing process must also consider the neurology and biome­ chanics of these joints, and their relationships with IVDs, muscle and other extra-articular structures.

zygapophysial joi nts spine vertebral col u m n development morpho logy joint ca psu le

l iga ments i njury tra u m a biomecha n ics i n nervation manual thera py

Boyle J W W, Milne N, Singer K P 2002 Influence of age on cervicothoracic spinal curvature: postural implications. Clinical Biomechanics 17: 361-367 Boyle J J W, Woodland P, Singer K P 2004 Patterns of fracture

/ dislocation at the cervicothoracic junctional region: an Australian
perspective. Spine (forthcoming) Bullough P G, Boachie-Adjei 0 1988 Atlas of spinal disorders. Lippincott, Philadelphia Burkus J 1988 Cervical facet asymmetry simulating facet dislocation. Spine 13: 118-120 Christ B, Wilting J 1992 From somites to vertebral column. Annals of Anatomy 174: 23-32 Cihak R 1970 Variations of lumbosacral joints and their morphogenesis. Acta Universitatis Carolinae Medica 16: 145-165 Cihak R 1981 Die Morphologie und Entwicklung der Wirbelbo­ gengelenke. Die Wirbelsaule in Forschung und Praxis 87: 13-28 Cyron B, Hutton W 1980 Articular tropism and stability of the lumbar spine. Spine 5: 1 68-172 Davis P 1955 The thoraco-Iumbar mortice joint. Journal of Anatomy 89: 370-377 Davis P R 1959 The medial inclination of the human thoracic intervertebral articular facets. Journal of Anatomy 93: 68-74 Farfan H 1969 The effects of torsion on the intervertebral joints. Canadian Journal of Surgery 12: 336-341 Farfan H 1973 Mechanical disorders of the low back. Lea and Febiger, Philadelphia Farfan H 1983 The torsional injury of the lumbar spine. Spine 8: 53 Farfan H F 1995 Form and function of the musculoskeletal system as revealed by mathematical analysis of the lumbar spine. Spine 20: 1462-1474 Farfan H F, Sullivan J D 1967 The relation of facet orientation to intervertebral disc failure. Canadian Journal of Surgery 10: 179-185 Farfan H, Huberdeau R, Dubow H 1972 Lumbar intervertebral disc degeneration. The influence of geometrical features on the pattern



of disc degeneration: a post mortem study. Journal of Bone and Joint Surgery 54-B: 492-51 0 Giles L , Harvey A 1987 Immunohistochemical demonstration of nociceptors in the capsule and synovial folds of human zygapophyseal joints. British Journal of Rheumatology 26: 362-364 Gregersen G, Lucas D 1967 An in vivo study of the axial rotation of the human thoracolumbar spine. Journal of Bone and Joint Surgery 49-A: 247-262 Grieve G 1981 Common vertebral joint problems. Churchill Livingstone, Edinburgh Groen G J, Stolker R J 2000 Thoracic neural anatomy. In: Giles L, Singer K P (eds) Clinical anatomy and management of thoracic spine pain. Butterworth Heinemann, Oxford, pp 114-142 Groen G J, Baljet B, Drukker J 1990 Nerves and nerve plexuses of the human vertebral column. American Journal of Anatomy 188: 282-296 Humphry G M 1858 A treatise on the human skeleton. Macmillan, London Kenesi C, Lesur E 1985 Orientation of the articular processes at L4, L5 and Sl: possible role in pathology of the intervertebral disc. Anatomica Clinica 7: 43-47 Kippers V, Parker A W 1985 Electromyographic studies of erectores spinae: symmetrical postures and sagittal trunk motion. Australian Journal of Physiotherapy 31: 95-105 Lewin T, Moffett B, Viidik A 1962 The morphology of the lumbar synovial intervertebral joints. Acta Morphologica Neerlando Scandinavica 4: 299-319 Lutz G 1967 Die Entwicklung der kleinen Wirbelgelenke. Zeitschrift fur Orthopadie und ihre Grenzgebiete 104: 19-28 McCall I W, Park W M, O'Brien J P 1979 Induced pain referral from posterior lumbar elements in normal subjects. Spine 4: 441-446 Macintosh J, Valencia F, Bogduk N, Munro R 1986 The morphology of the human lumbar multifidus. Clinical Biomechanics 1: 196-204 Maigne J Y, Ayral X, Guerin-Surville H 1992 Frequency and size of ossifications in the caudal attachments of the ligamentum flavum of the thoracic spine: role of rotatory strains in their development. Surgical and Radiologic Anatomy 14: 119-124 Malmivaara A, Videman T, Kuosma E, Troup J D G 1987 Facet joint orientation, facet and costovertebral joint osteoarthrosis, disc degeneration, vertebral body osteophytosis and Schmorl's nodes in the thoracolumbar junctional region of cadaveric spines. Spine 12: 458-463 Markolf K L 1972 Deformation of the thoracolumbar intervertebral joints in response to external loads. Journal of Bone and Joint Surgery 54A: 511-533 Med M 1972 Articulations of the thoracic vertebrae and their variability. Folia Morphologica 20: 212-215 Med M 1973 Articulations of the cervical spine and their variability. Folia Morphologica 21: 324-327 Med M 1977 Prenatal development of thoracic intervertebral articulations. Folia Morphologica 25: 175-177 Med M 1980 Prenatal development of intervertebral articulation in man and its association with ventrodorsal curvature of the spine. Folia Morphologica 28: 264-267 Mercer S, Bogduk N 1993 Intra-articular inclusions of the cervical synovial joints. British Journal of Rheumatology 32: 705-710 Milne N 1991 The role of zygapophysial joint orientation and uncinate processes in controlling motion in the cervical spine. Journal of Anatomy 178: 1 89-201 Milne N 1993a Comparative anatomy and function of the uncinate processes of cervical vertebrae in humans and other mammals. PhD thesis, University of Western Australia, Perth Milne N 1993b Composite motion in cervical disc segments. Clinical Biomechanics 8: 1 93-202 Mooney V, Robertson J 1976 The facet syndrome. Clinical Orthopaedics 115: 149-156

Odgers P 1933 The lumbar and lumbo-sacral diarthrodial joints. Journal of Anatomy 67: 301-317 O'Rahilly R, Muller F, Meyer D B 1980 The human vertebral column at the end of the embryonic period proper. 1. The column as a whole. Journal of Anatomy 131: 565-575 Pal G, Routal R 1986 A study of weight transmission through the cervical and upper thoracic regions of the vertebral column in man. Journal of Anatomy 148: 245-261 Pal G, Routal R 1987 Transmission of weight through the lower thoracic and lumbar regions of the vertebral column in man. Journal of Anatomy 152: 93-105 Pearcy M J 1997 Biomechanics of the lumbosacral spine. In: Giles L, Singer K P (eds) Clinical anatomy and management of low back pain. Butterworth Heinemann, Oxford, pp 165-172 Pearcy M J, Hindle R J 1991 Axial rotation of lumbar intervertebral joints in forward flexion. Proceedings of the Institute of Mechanical Engineers 205: 205-209 Penning L, Wilmink J T 1987 Rotation of the cervical spine. Spine 12: 732-738 Ponseti I V 1995 Differences in ligamenta flava among some mammals. Iowa Orthopaedic Journal 15: 141-146 Prescher A 1997 The craniovertebral junction in man, the osseous variations, their significance and differential diagnosis. Annals of Anatomy 179: 1-19 Putti V 1927 New conceptions on the pathogenesis of sciatic pain. Lancet 2: 53-60 Putz R 1985 The functional morphology of the superior articular processes of the lumbar vertebrae. Journal of Anatomy 143: 181-187 Reichmann S 1971 The postnatal development of form and orientation of the lumbar intervertebral joint surfaces. Zeitschrift fur Anatomie Entwicklungsgeschichte 133: 102-123 Rickenbacher J, Landolt A M, Theiler K 1985 Applied anatomy of the back. Springer-Verlag, Berlin pp 30, 31 Rolander S D 1966 Motion of the lumbar spine with special reference to the stabilizing effect of posterior fusion. Acta Orthopedica Scandinavia 90 (Supp!.): 1-144 Saada J, Song S, Breidahl W H 2000 Developmental anomalies of the thoracic region. In: Giles L, Singer K P (eds) Clinical anatomy and management of thoracic spine pain. Butterworth Heinemann, Oxford, pp 83-99 Schmorl G, Junghanns H 1971 The human spine in health and disease. Grune and Stratton, New York Singer K P 1989 The thoracolumbar mortice joint: radiological and histological observations. Clinical Biomechanics 4: 137-143 Singer K P 2000 Pathology of the thoracic spine. In: Giles L, Singer K P (eds) Clinical anatomy and management of thoracic spine pain. Butterworth Heinemann, Oxford, pp 63-82 Singer K P, Breidahl P D 1990 Accessory ossification centres at the thoracolumbar junction. Surgical and Radiologic Anatomy 12: 53-58 Singer K P, Giles L G F 1990 Manual therapy considerations at the thoracolumbar junction: an anatomical and functional perspective. Journal of Manipulative and Physiological Therapeutics 13: 83-88 Singer K P, Breidahl P D, Day R E 1988 Variations in zygap9physeal orientation and level of transition at the thoracolumbar junction: a preliminary CT survey. Surgical and Radiologic Anatomy 10: 291-295 Singer K P, Breidahl P D, Day R E 1989a Posterior element variation at the thoracolumbar transition: a morphometric study using computed tomography. Clinical Biomechanics 4: 80-86 Singer K P, Day R E, Breidahl P D 1989b In vivo axial rotation at the thoracolumbar junction: an investigation using low dose CT in healthy male volunteers. Clinical Biomechanics 4: 145-150 Singer K P, Willen J, Breidahl P D, Day R E 1989. The influence of zygapophyseal joint orientation on spinal injuries at the thoracolumbar junction. Surgical and Radiologic Anatomy 11: 233-239

Comparative anatomy of the zygapophysial joints


Singer K P, Giles L G F, Day R E 1990 Intra-articular synovial folds of the thoracolumbar junction zygapophyseal joints. Anatomical Recon;j 226: 147-152 Singer K 'p, Svansson G, Day R E, Breidahl W H, Horrex A 2001 The utility of diagnosing lumbar rotational instability from twist CT scans. Journal of Musculoskeletal Research 5: 45-51 Stokes I A F 1988 Mechanical function of facet jOints in the lumbar spine. Clinical Biomechanics 3: 101-105 Sward L, Hellstrom M, Jacobsson B, Nyman R, Peterson L 1991 Disc degeneration and associated abnormalities of the spine in elite gymnasts: MR1 study. Spine 16: 437-443 Tan C I, Kent G N, Randall A G, Edmondston J, Singer K P 2003 Age­ related changes in collagen, pyridinoline and deoxypyridinoline in

normal human thoracic intervertebral discs. Journal of Gerontology: Biological Sciences 58(5B): 387-393 Taylor J R, Twomey L T 1986 Age changes in lumbar zygapophyseal joints: observations on structure and function. Spine 1 1 : 739-745 Tondury G 1971 Functional anatomy of the small joints of the spine. Annales de Medecine Physique 15: 173-191 Tondury G 1972 Anatomie fonctionelle des petites articulations de rachis. Annales de Medecine Physique 15: 1 73-191 Whitney C 1926 Asymmetry of vertebral articular processes and facets. American Journal of Physical Anthropology 9: 451-455 Yang K, King A 1984 Mechanism of facet load transmission as a hypothesis for low back pain. Spine 9: 559-565



Chapter 4

Kinematics of the spine
S. Mercer



31 32 Lower cervical spine 33 Thoracic spine 34 Lumbar spine 34
Atlanto-occipital joint Atlanto-axial joint

understanding of movement of the spine is essential to comprehension of its normal function. One of the most fun­ damental parameters of spinal motion is spinal range of motion,which is often used as an index of spinal function. The normative data against which impairment ratings are made have been collected from cadavers and from living individuals using a variety of techniques including external devices or radiography. Shortcomings of this normative data lie in the lack of generalizability of subjects, lack of reliability of the measuring instruments and lack of validity between external instruments and radiological techniques. In addition cadaver studies cannot be generalized to living individuals as the motion and resistance provided by mus­ cles have been removed. But most importantly,measures of global range of motion do not reveal what is happening inside the neck or trunk. Recognition of the shortcomings of these global range of motion studies led to studies examining segmental motion. These technically more difficult investigations have also examined cadavers and living individuals with external devices, radiographs and computed tomography ( CT). They have provided data regarding segmental motion including patterns of coupled motion. The purpose of this chapter is to describe spinal kinematics in terms of seg­ mental motion, highlighting the clinically relevant gaps in our knowledge.


The deep atlantaI sockets of the atlas are designed to cra­ dle the occiput and transmit forces from the head to the cervical spine. This design facilitates flexion and extension but impedes other movements ( Mercer & Bogduk 2001). In living individuals the average mean motion is about 14-15 degrees (Table 4.1), although Fielding (1957) reported a much higher value of 35 degrees. However,the variability in range of motion in normal subjects is large, being 0-22 degrees (Kottke & Mundale 1959) or 0-25 degrees (Brocher 1955). Furthermore, Lind et al (1989)



reported a mean of 14 degrees with a standard deviation of 15 degrees in normal subjects. Such wide variations in reported normal flexion and extension range of motion must be taken into account when making decisions about what constitutes normal or abnormal movement at the atlanto-occipital joint. These variations could be due to differences in the way in which the occipital flexion and extension movements were performed or to the paradoxi­ cal motion of the atlas that different postural strategies may induce (Bogduk & Mercer 2000). Other more detailed information regarding the kinemat­ ics of the atlanto-occipital joints comes from studies on cadaveric material (Werne 1958, Worth 1985, Worth & Selvik 1986). Werne (1958) measured 13 degrees of flex­ ion-extension and 0 degrees of axial rotation, although he was able to measure 8 degrees of axial rotation when the movement was forced. A more precise radiographic study described the mean range (SO) of flexion-extension at 18.6 degrees (0.6),axial rotation 3.4 degrees (0.4) and lateral flex­ ion 3.9 degrees (0.6) (Worth 1985,Worth & Selvik 1986). During flexion-extension negligible motion was observed in the other planes; however,during axial rotation 1.5 degrees of extension and 2.7 degrees of lateral flexion were recorded ( Worth 1985, Worth & Selvik 1986). Therefore in cadavers axial rotation was artificially created through a combination of extension and lateral flexion. This pattern of coupling should not necessarily be accepted as the normal pattern of coupling as it could be the result of how and when the axial torque was applied to the cadavers (Bogduk & Mercer 2000). We do not know whether this is the pattern of coupling that occurs in vivo when muscles are active or whether posture would affect such patterns of coupled motion. When inducing lateral flexion, Worth & Selvik (1986) noted that this movement could be coupled with flexion,extension or axial rotation,with the pattern of coupling being dependent on the shape of the atlantal sock­ ets. As individual anatomical variation may therefore influ­ ence the pattern of coupling and as there is a dearth of studies examining atlanto-occipital joint motion,particular rules for patterns of defined coupled motion are not sup­ ported by the current literature.
Table Table


Normal ranges of motion of in vivo flexion-

extension at the atlanto-occipital joint



Range of motion (degrees) Range SD

Brocher 1955 Lewit Et Krausova 1963 Markuske 1971 Fielding 1957 Kottke Et Mundale 1959 Lind et al 1981

14.3 15.0 14.5 35.0 14.0


0-22 15


Studies examining range of motion at the atlanto-axial joints in cadavers report 10 degrees of flexion-extension and 47 degrees of axial rotation (Werne 1958),and about 5 degrees of lateral flexion ( Oankmeijer & Rethmeier 1943). A more recent study using CT scanning observed 32 degrees (50,10) of axial rotation to either side (Dvorak et aI 1987a). In living individuals the reported range of flexion­ extension motion is highly variable,varying between 2 and 1 8 degrees (Table 4.2). Due to the difficulties in accurately determining from plain X-rays the range of axial rotation, most studies have only examined flexion-extension at the atlanto-axial joints. Mimura et al (1989) used biplanar radiography to more accurately examine atlanto-axial joint motion. The total range of axial rotation ( left to right) of the occiput relative to C2 was 75.2 degrees ( SO, 11.8). This axial rotation was accompanied by 14 degrees ( SO, 6) of extension and 24 degrees (SO,6) of contralateral lateral flexion,although the authors reported that in some cases flexion would accom­ pany the axial rotation rather than extension. This variabil­ ity in coupling occurs because of the passive nature of the kinematics of the atlas ( Mercer & Bogduk 2001). Whether the atlas flexes or extends during axial rotation depends on the geometry of the atlanto-axial joints and the precise direction of any forces acting through the atlas from the head (Bogduk & Mercer 2000).


Normal ranges of motion at the atlanto-axial joint in living individuals


Ranges of motion (degrees) Flexion-extension Axial rotation One side Total

Brocher 1955 Kottke Et Mundale 1959 Lewit Et Krausova 1983 Markuske 1971 Lind et al 1989 Fielding 1957 Hohl Et Baker 1964

18 (2-16) 11 16 21 13 ( +/-5) 15 (10-15)

90 30

Kinematics of the spine


In normal living subjects imaged via CT scanning a mean of 43 degrees (SO, 5.5) of axial rotation was measured to each side at Cl-2 with a left-right asymmetry of 2.8 (50,2) (Dvorak et aI 1987b). This finding led these authors to sug­ gest that 56 degrees is an upper limit of normal axial rotation.

The general pattern of segmental motion during flexion and extensi.on of the cervical spine has been described by van Mameren (1988). Flexion may be divided into three sequential phases. The initial phase begins in the lower cer­ vical spine (C4-7) where C6-7 makes its maximum contri­ bution followed by the C5-6 segment and then by C4-5. Motion in the second phase occurs initially at CO-2 fol­ lowed by C2-3 and C3-4,the order of contribution of C2-3 and C3-4 being variable. During this phase slight extension occurs at C6-7 and in some individuals at C5-6. The third phase of motion occurs again at the lower cervical spine (C4-7) initially, with the C4-5 segment followed by C5-6 then C6-7 s2gment. Flexion in normal subjects is therefore initiated and terminated by C6-7,never by the mid-cervical segments. The CO-2 and C2-3, C3-4 segments contribute maximally during the middle phase of motion, but in a variable sequence (Bogduk & Mercer 2000). Extension may also be divided into three phases (van Mameren 1988). The first phase is initiated in the lower cer­ vical spine (C4-7) with no regular pattern to the sequence of segmental motion. In the middle phase,motion occurs at <20-2 and at C2-4 with the order of contribution being quite variable between C2 and C4. The third phase is character­ ized by a second contribution from the lower segments (C4-7) in which the individual segments move in a regular order C4-5, C5-6, then C6-7. Meanwhile motion of CO-2 attains its maximum range. This pattern of motion during flexion and extension was shown to be reproducible (van Mameren 1988). Although many studies have been undertaken to deter­ mine normal ranges of segmental motion during flexion and extension of the cervical spine, few have included mean range and standard deviation of this motion. Two early studies (Aho et al 1955,Bhalla & Simmons 1969) pro-

vided raw data so that means and standard deviations can be calculated (Bogduk & Mercer 2000), while two more recent studies (Dvorak et al 1988,Lind et al 1989) also afford more meaningful normative data for clinicians (Table 4.3). However,only Lind et al (1989) and Dvorak et al (1988) also report the inter-observer error of their measurement tech­ nique, therefore providing the most reliable normative data. Examination of Table 4.3 reveals the largest range of flexion-extension motion at the C4-5 and C5-6 segments. The work of van Mameren (van Mameren et al 1990) has highlighted the difficulties of using normative segmental motion data for clinical purposes. This study demonstrated that in normal subjects the total range of motion of the neck is not the arithmetical sum of its intersegmental ranges of motion. Further,segmental range of motion differs accord­ ing to whether the motion is performed from flexion to extension or from extension to flexion resulting in differ­ ences of 10-30 degrees in total range of cervical motion. Finally, the ranges of motion are not stable over time (Bogduk & Mercer 2000). The clinical implication of this study is that normal motion must be considered as a fluc­ tuating range of values and not as a single value. At the segmental level,flexion is a movement composed of anterior sagittal rotation and anterior translation. The extent of coupling between the rotation and translation is determined by the height of the superior articular process (Nowitzke et al 1994). As the superior articular processes are shorter at higher cervical levels these segments exhibit relatively greater amplitude of translation, while at lower levels the taller superior articular processes impede transla­ tion resulting in a greater ratio of rotation to translation. Using CT scanning in the conventional horizontal plane Penning & Wilmink (1987) determined the mean and ranges of axial rotation at each level within the cervical spine (Table 4.4). Due to the structure of the cervical spine, axial rotation in the horizontal plane is, however, inescapably coupled with ipsilateral lateral flexion. Consequently when axial rotation has been examined by CT scanning in the horizontal plane the ranges of axial rota­ tion computed have been confounded by movement of the plane of view. Therefore the normal values provided in Table 4.6 are only an imprecise estimate of the range of seg­ mental axial rotation within the cervical spine.



Mean and standard deviation in degrees for segmental motion during

cervical flexion and extension


Mean values and (SD) of flexion and extension motion C2-3 C3-4 15 15 14 15 (7) (2) (6) (3) C4-5 22 23 16 19 (4) (1) (6) (4) C5-6 28 19 15 20 (4) (1) (8) (4) C6-7 15 (4) 18 (3) 11 (7) 19 (4)

Aho et al 1955 Bhalla Et Simmons 1969 Lind et al 1981 Dvorak et ai' 1988

12 9 10 10

(5) (1) (4) (3)





Mean and range of axial rotation of cervical

motion segments (based on Penning Et Wilmink


(2000) calculated the range . of axial rotation in the plane of the zygapophysial joints to be about 8 degrees for a range of 6 degrees of horizontal rotation.


Range of motion (degrees) Mean Range -2-5 29-46 0-10 3-10 1- 12 2- 12 2-10 -2-7

Occ-Cl Cl-2 C2-3 C3-4 C4-5 C5-6 C6-7 C7-Tl

1.0 40.5 3.0 6.5 6.B 6.9 2. 1 2. 1

Mimura et al (1989) subsequently used trigonometric reconstruction of motion recorded via biplanar radiography to provide a more valid measurement of axial rotation and coupled motion within the cervical spine. The normal ranges of motion are provided in Table 4.5. Coupling of flex­ ion with axial rotation was observed at C� and C6--7 while extension was seen above C4--5. Ipsilateral lateral flexion coupled with axial rotation caudal from C3 to C4. However, what must be kept in mind are the large variations in nor­ mal motion as expressed in the standard deviations being much larger than the mean values of motion (Table 4.7). If motion in the cervical spine is considered in terms of form and function rather than following the traditional con­ vention of planes the pure movements available are flexion--extension around the coronal axis and axial rotation around an axis perpendicular to the facets of the zygapophysial joints (Bogduk & Mercer 2000). Axial rota­ tion in the plane of the facet rather than in the horizontal plane results in the inferior articular processes gliding freely across the superior articular facets of the vertebra below while the superior vertebral body rotates within the concavity formed by the uncinate processes below (Bogduk & Mercer 2000, Mercer & Bogduk 2001). Bogduk & Mercer

The few studies describing motion in the thoracic spine have been undertaken on cadaveric material or on living individuals using external devices that have not been vali­ dated with regard to segmental motion. Consequently the research literature at present provides limited clinically useful data. White (1969) measured range of motion in three planes on cadaveric thoracic spines (Table 4.6). Examination of Table 4.8 reveals the pattern of cephalocaudal variation in range of motion. There is a tendency for greater flexion--extension at lower levels of the thoracic spine, no particular pattern is evident for lateral flexion, while less axial rotation is found at lower vertebral levels. Studies which have examined coupling in the thoracic spine have reported that lateral flexion may be coupled with ipsilateral (Gregersen & Lucas 1967, White 1969) or equally ipsilateral or contralateral axial rotation in the upper region (Willems et al 1996), while in the middle or lower regions of the thoracic spine lateral flexion may be coupled ipsilaterally or contralaterally with axial rotation· (White 1969) or predominantly ipsilaterally (Gregerson & Lucas 1967, Willems et aI 1996). The differences in meth­ ods and subjects in each of these few studies influence the variability in reported data. At present there exist insuffi­ cient data to support models of specific patterns of cou­ pling within the thoracic spine.

Flexion of the lumbar spine from erect standing involves an unfolding or straightening of the lumbar lordosis followed




(SO) of segmental motion in degrees 1 969)
Lateral flexion 6.0 (3.B) 4.B(3.4) 3.7(2.1) 5.0 (5.1) 5.3(2.9) 4.2(5.5) 4. 1 (3.0) 3.7(l.B) 4.4(loB) 4.4(2.3) 3.7(3.6) Axial . rotation 4.0 ( 1.9) 5. 1(3.4) 3.9( 1.4) 5.0(2.1) 4. 1 ( 1.3) 4.3( 1.7) 5.5 J2.1) 3.2(1.7) 3.4 (l.B) 2.6(0.3)

during total flexion-extension, lateral flexion and axial rotation (based on White Table ion

4.5 Normal ranges of axial rotation and coupled flex(+) and extension (-), ipsilateral (+) and contralateral (-) (SO) (based on Mimura et al 1989)
Coupled movement Axial rotation Flexion/extension - 1 4(6) 0(3) -3(5) -2(4) 2 (3) 3 (3) Lateral flexion -2(6) -2(B) 6(7) 6(7) 4(B) 3 (7)

lateral flexion associated with right or left axial rotation, mean degrees and

Vertebral level Tl-2 T2-3 T3-4 T4-5 T5-6 T6-7 T7-B TB-9 T9-10 TlO-ll Tll- 12

Flexionextension 2.B(O.B) 2.6(O.B) 2.3( 1.7) 1.B (0.9) 2.6( 1.0) 2.3 ( 1.3) 3.3 ( 1.7) 3.2 (l.B) 3. 1 ( 1.4) 3.9 (2.7) 6.5(4.7)


Occ-C2 C2-3 C3-4 C4-5 C5-6 C6-7

75 ( 12) 7(6) 6 (5) 4 (6) 5(4) 6 (3)

Kinematics of the spine


by at most a small reversal of the lordotic curve. Kanayama et al (1996) reported that motion usually begins in the upper lumbar region followed by the rest of the lumbar spine or else all segments move at the same time. Extension is essentially the reverse of flexion. However, Gatton & Pearcy (1999) observed no consistency in sequencing of seg­ mental motion during flexion. Further, they found that nor­ mal subjects may use any of a variety of movement strategies to achieve flexion of the lumbar spine. When per­ forming lumbar flexion twice some subjects used different movement. sequences each time. While examining trunk extension from flexion Okawa et al (1998) also found no particular pattern of regional motion to occur. The ranges for segmental motion during flexion and extension are shown in Table 4.7. These data come from the studies of Pearcy et al (1984) and Pearcy & Tibrewal (1984) who examined males 25-36 years of age. When considering motion in the sagittal plane all lumbar joints have a similar total range of motion; however, the highest and lowest joints have relatively greater range of extension while the middle joints exhibit a relatively greater range of flexion. However, what must be remembered is that these results may not necessarily be generalized to males of other ages or to females. At the segmental level lumbar flexion occurs through anterior sagittal rotation and simultaneous anterior sagit­ tal translation of each vertebra (Okawa et al 1998) while during extension posterior sagittal rotation and posterior translation occur. During flexion each vertebra rotates and translates from the backward tilted position adopted with the assumption of the lordosis to a neutral position where the superior and inferior surfaces of adjacent vertebrae become parallel. Additional movement in the upper lum­ bar region allows a reversal of the lordotic curve to occur with the upper vertebrae rotating further forwards



Ranges of segmental motion during flexion,

extension and flexion and extension

Mean range of motion in degrees (SO) Segment Ll-2 L2-3 L3-4 L4-5 L5-S1 Flexion 8 (5) 10 (2) 12(1) 13(4) 9 (6) Extension 5 (2) 3 (2) 1(1) 2 (1) 5 (4) Flexion and extension 13 (5) 13 (2) 13 (2) 16(4) 14 (5)

through anterior compression of the intervertebral discs (Bogduk 1997). The ranges of motion for rotation and translation in each plane during flexion and extension are depicted in Table 4.8. For each level the ranges of the sagittal plane motion during flexion lie between 8 and 13 degrees of rotation and 1-3 mm of translation. In addition these movements are regularly associated with 1 degree of coronal and axial rota­ tion and very small and variable amounts of lateral and vertical translation. During extension 1-5 degrees of sagit­ tal plane rotation is coupled with 1 degree of translation while very small amounts of coronal and axial rotation in association with negligible amounts of vertical and lateral translation occur (Table 4.8). These very small ranges avail­ able in conjunction with the large amount of variation do not confirm specific patterns of coupled motion which have been postulated in the clinical literature. During axial rotation the superior vertebrae will initially rotate about an axis located in the posterior annulus fibro­ sus (Cossette et aI1971). After about 3 degrees of axial rota­ tion the zygapophysial joint contralateral to the direction of rotation impacts. Any further axial rotation occurs around



Motion coupled with flexion and extension (based on Pearcy et al.


Level Mean (SO) rotations (degrees) Sagittal

Coupled motion Mean (SO) translations (mm) Sagittal 3 (1) 2 (1) 2(1) 2 (1) 1(1) Coronal 0(1) 1(1) 1(1) 0(1) 0(1) Axial 1(1) 1(1) 0(1) 0(1) 1(1)

Coronal 1(1) 1(1) 1(1) 2 (1) 1(1)

Axial 1(1) 1(1) 1(1) 1(1) 1(1)

Ll L2 L3 L4 L5

8 (5) 10(2) 12(1) 13(4) 9(6)

L1 L2 L3 L4 L5

5(1) 3(1) 1(1) 2(1) 5(1)

0(1) 0(1) 1(1) 1(1) 1(1)

1(1) 1(1) 0(1) 1(1) 1(1)

1(1) 1(1) 1(1) 1(1) 1(1)

1(1) 0(1) 1(1) 0(1) 1 (1)

0(1) 0(1) 0(1) 1(1) 0(1)



an axis that has now migrated to the impacted zygapophysial joint and so further rotation is at the expense of the intervertebral disc. Ranges of axial rotation studied via biplanar radiogra­ phy in 24-36-year-old males are depicted in Table 4.9 (Pearcy & Tribrewal 1 984, Pearcy et al 1 984). The small ranges of axial rotation available would appear to be s ��­ lar at all levels and all fall within the 3 degree safe limIt described in the biomechanical literature. Axial rotation tends to couple with contralateral lateral flexion at upper lumbar levels but with ipsilateral lateral flexion at L5/Sl. However,examination of Table 4.1 0 high­ lights the very small amounts of mean motion being dis­ cussed while the variation in size and direction of motion is large. Flexion or extension may couple with axial rotation resulting in the mean magnitude of coupled motion in the sagittal plane being zero. Lateral flexion is a complex and highly variable move­ ment. It involves lateral bending and rotatory movements of the interbody joints and a variety of movements at the zygapophysial joints (Bogduk 1 997). Biplanar radiography in young male adults reveals the smaller range of lateral flexion at lower levels of the lumbar spine when compared with upper segments (Table 4. 1 ) reflecting the bony and 1 ligamentous anatomy of the L5 and to a lesser extent the L4 vertebrae (Pearcy & Tibrewal1 984, Pearcy et aI1 984).



Ranges of segmental motion in degrees during

lateral flexion (based on Pearcy et al

1 984, Pearcy

8: Tibrewal

1 984)
Mean range of motion Segment L 1-2 L2-3 L3-4 L4-5 L5-S 1 Left 5 5 5 3 0 Right 6 6 6 5 2


Ranges of segmental motion in degrees during

axial rotation (based on Pearcy et al

1 984, Pearcy

8: Tribrewal

Mean range of motion in degrees Segment L 1-2 L2-3 L3-4 L4-5 L5-S 1 Left Right

The general pattern of coupled motion is for lateral flex­ ion to be associated with contralateral axial rotation at upper lumbar levels but ipsilateral axial rotation at LS/S 1 (Table 4.1 2). Lateral flexion may be accompanied by either flexion or extension but as extension occurs more fre­ quently and at a larger magnitude, the results of this study suggest that lateral flexion is usually accompanied by extension (Pearcy & Tibrewal1 984). The work by Pearcy and colleagues highlights the fact that the patterns of coupled motion in the lumbar spine are only general patterns that are subject to a high degree of variability both between subjects and between segments within the one individual. A normal subject may exhibit the general pattern of coupling at one level but may exhibit reverse coupling at other levels ( Pearcy 1 985). Consequently there is at present little evidence for strict rules of coupled motion that determine whether an indi­ vidual has abnormal ranges or directions of coupling in the lumbar spine.

2 2

KEYWORDS kinematics spinal range of motion atlanto-occipital joint atlanto-axial joint



Coupled movements of lumbar spine rotation (based on Pearcy 8: Tibrewal

1 984)


Axial rotation (degrees) Mean Range (-2- 1) (- 1-1) (-2- 1) (-1-1) (-3-1) (0-1) (-2- 1) (0- 1) (-2- 1) (-2-1)

Lateral flexion (degrees) Mean 3 -3 4 -3 3 -3 1 -2 -2 Range (-1-5) (-7 to- 1) ( 1-9) (-5-0) ( 1-6) (-6-0) (-7-0) (-5-1) (-7-0) (0-2)

Flexion/extension (degrees) Mean 0 0 0 0 0 0 0 0 0 0 Range ( 3 3) (-4-4) (-2-2) (-4-4) (-2-2) (-3-2) (-9-5) (-7-2) (-5-3) (-5-3)

R L1 L L1 R L2 L L2 R L3 L L3 R L4 L L4 R L5 L L5

-1 1 -1 -1 2 -1 2 -1 0

Kinematics of the spine



4.1 2

Coupled movements of the lumbar spine during lateral flexion (based on Pearcy Et Tibrewal



Lateral flexion (degrees) Mean Range (-8 to -2) (4-10) (-8 to -4) (2-10) (-11-2) (-3-8)

Axial rotation (degrees) Mean 0 0 -1 -1 -1 0 -2 Range (-3- 1) (-2- 1) H - 1) (-3- 1) H - 1) (-4-1) (0- 1) (-4-1) H - 1) (-3-1)

Flexion/extension (degrees) Mean -2 -2 -1 -3 -1 -2 0 -1 2 0 Range (-5- 1) (-9-0) (-3- 1) (-4 to- 1) (-3- 1) (-4-3) H-4) (-4-2) (-3-8)

Rl 1 Ll1 R L2 L L2 R L3 L L3 RL4 L L4 RL5 L L5


-5 -5
5 -3 3 0 3



(-3-6) (-2-3) (-6-1)


Aho A, Vartianen 0, Salo 0 1955 Segmentary antero-posterio mobility of the cervical spine. Annales Medicinae Internae Fenniae 44: 287-299 Bhalla S K, SimmonsE H 1969 Normal ranges of intervertebral joint motion of the cervical spine. Canadian Journal of Surgery 12: 181-187 Bogduk N 1997 Clinical anatomy of the lumbar spine and sacrum, 3rd edn. Churchill Livingstone,Edinburgh Bogduk N, Mercer S R 2000 Biomechanics of the cervical spine. I: Normal kinematics. Clinical Biomechanics 15: 633--648 Brocher JEW 1955 Die occipito-cervical-gegend: eine diagnostiche pathogenetische studie. Georg Thieme Verlag, Stuttgart Cossette JW, Farfan H F, Robertson G H, Wells R V 1971 The instantaneous center of rotation of the third lumbar intervertebral joint. Journal of Biomechanics 4: 149-153 Dankmeijer J, Rethmeier B J 1943The lateral movement in the atlanto­ axial joints and its clinical significance. Acta Radiologica 24: 55---{;6 Dvorak J, Panjabi M, Gerber M, Wichmann W 1987a CT-functional diagnostics of the rotatory instability of upper cervical spine. 1: An experimental study on cadavers. Spine 12: 197-205 Dvorak J, Hayek J, Zehnder R 1987b CT-functional diagnostics of the rotatory instability of the upper cervical spine. 2: An evaluation on healthy adults and patients with suspected instability. Spine 12: 725--731 Dvorak J, Froehlich D, Penning L, Baumgartner H, Panjabi M M 1988 Functional radiographic diagnOSiS of the cervical spine: flexion/extension. Spine 13: 748-755 Fielding JW 1957 Cineradiography of the normal cervical spine. Journal of Bone and Joint Surgery 39: 1280-1288 Gatton M L, Pearcy M J 1999 Kinematics and movement sequencing during flexion of the lumbar spine. Clinical Biomechanics 14: 376-383 Gregersen G, Lucas D 1967 An in vivo study of the axial rotation of the human thoracolumbar spine. Journal of Bone and Joint Surgery 49A: 247-262 Hohl M, Baker H R 1964 The atlanto-axial joint. Journal of Bone and Joint Surgery 46A: 1739-1752 Kanayama M, Abumi K, Kaneda K, Tadano S, Ukai T 1996 Phase lag of the intersegmental motion in flexion-extension of the lumbar and lumbosacral spine: an in vivo study. Spine 21: 1416-1422 Kottke F J, Mundale M 0 1959 Range of mobility of the cervical spine. Archives of PhYSical Medicine and Rehabilitation 40: 379-382 Lewit K, Krausova L 1963 Messungen von vor-und ruckbeuge in den kopfgelenken. Fortschrift Rontgenst 99: 538-549 Lind B, Sihlbom H, Nordwall A, Malchau H 1989 Normal range of motion of the cervical spine. Archives Physical Medicine and Rehabilitation 70: 692---{;95 Markuske H 1971 Untersuchungen zur statik und dynamik der kindlichen halswirbelsaule: der aussagewert seitlicher rontgenaufnahmen: die wirbelsaule in forschung und praxis. Hippokrates, Stuttgart Mercer S R, Bogduk N 2001 The joints of the cervical vertebral column. Journal of Orthopaedic and Sports Physical Therapy 31: 174-182 Mimura M, Moriya H,Watanabe T, Takahashi K, Yamagata M, Tamaki T 1989 Three-dimensional motion analysis of the cervical spine with special reference to the axial rota tion. Spine 14: 1135-1139 Nowitzke A, Westaway M, Bogduk N 1994 Cervical zygapophyseal joints: geometrical parameters and relationship to cervical kinematics. Clinical Biomechanics 9: 342-348 Okawa A, Shinomiyaw K, Komori H, Muneta T, Arai Y, Nakai 0 1998 Dynamic motion study of the whole lumbar spine by videofluoroscopy. Spine 23: 1743-1749 Pearcy M J 1985 Stereo-radiography of lumbar spine motion. Acta Orthopaedica Scandinavica 212 (Supp!.): 1-41 Pearcy M J, Tibrewal S B 1984 Axial rotation and lateral bending in the normal lumbar spine measured by three-dimensional radiography. Spine 9: 582-587 Pearcy M, Portek I, Shepherd J 1984 Three-dimensional X-ray analysis of normal movement in the lumbar spine. Spine 9: 294-297 Penning L,Wilmink J T 1987 Rotation of the cervical spine: a CT study in normal subjects. Spine 12: 732-738 van Mameren H 1988 Motion patterns in the cervical spine. Thesis, University of Limberg, Maastricht van Mameren H, Drukker J, Sanches H, Beursgens J 1990 Cervical spine motion in the sagittal plane. I: Range of motion of actually performed movements, an X-ray cinematographic study.European Journal of Morphology 28: 47-68 Werne S 1958 The possibilities of movement in the craniovertebral joints. Acta Orthopaedica Scandinavica 28: 165-173 W hite A A 1969 Analysis of the mechanics of the thoracic spine in man. Acta Orthopaedica Scandinavica 127 (Supp!.): 1-105 Willems J M, Jull G A, Ng J K-F 1996 An in vivo study of the primary and coupled rotations of the thoracic spine. Clinical Biomechanics 11: 311-316 Worth D 1985 Cervical spine kinematics. PhD Thesis, Flinders University of South Australia Worth D, Selvik G 1986 Movements of the craniovertebral joints. In: Grieve G (ed) Modern manual therapy of the vertebral column. Churchill Livingstone,Edinburgh





Chemistry of the intervertebral disc In relation to functional requirements
J. P. Urban. S. Roberts

39 39

Extracellular influences on disc cell metabolism Nutrient levels

49 49 49 50 50 50

Gross structure of the disc
The nucleus pulposus The anr1ulus fibrosus

Growth factors and cytokines Mechanical stress In vitro studies

39 40 The cartilage end-plate .40
The constituents of the disc
Proteoglycans Collagen

In vivo responses to load


40 40 42 42 42



Disc proteoglycans

The intervertebral disc has a prominent role in the structure and function of the spine. It is able to transmit load and act as a joint. Although its mechanical behaviour in compres­ sion, extension, torsion and bending has been extensively

Collagen types in disc Other matrix constituents Proteinases Cells Water

Collagen organization in the disc


43 43 43 44 44 44 45 45 44

Nerves and blood vessels

investigated, little is yet known of how the composition and structure of the disc influence its mechanical behaviour. In this chapter, current ideas on the relationship between disc mechanical function and its chemical composition will be reviewed.

Disc diseases

Spinal deformities Disc herniation Disc degeneration
A structural model

Function of the constituents of the disc

The biophysical function of proteoglycans and collagen

The intervertebral disc is generally considered to consist of

45 45 45 45

The osmotic pressure of proteoglycan solutions Swelling pressure of the disc Hydraulic permeability Swelling pressure and tissue fluid content

two distinct regions: the outer, firm, banded annulus fibro­ sus and the inner, soft, gelatinous nucleus pulposus. The cartilaginous end-plates are interposed between the bony vertebral bodies and the disc itself. Figure 5.1 illustrates the regions of the disc. Some useful reviews of disc structure and the changes found with age and with degeneration have been published (Coventry et al 1945, Peacock 1951, Buckwalter 1995, Urban & Roberts 1995).

46 46

Solute partitions and transport

Mechanical behaviour of the disc in relation to its composition

46 46 46 47

Loads on the lumbar spine

The deformation of the disc under load

Cell energy metabolism and disc nutrition
Energy metabolism Disc nutrition


The nucleus pulposus
The nucleus occupies the central region of the disc. Its com­ position and appearance change markedly throughout life. In children it is highly hydrated, being 85-90% water, and is white and translucent. There is a clear demarcation

48 49

Cell metabolism and matrix turnover
In vivo measurements





c D



Schematic view of the disc and vertebral body



Schematic view of the disc extracellular matrix show­

showing (a) vertebral body, (b) cartilaginous end-plate, (c) nucleus pulposus, (d) annulus fibrosus.

ing details of the aggrecan molecule.

fibres embedded in a proteoglycan- water gel (Fig. 5.2). between it and the surrounding annulus fibrosus. In adults the hydration drops markedly and, as the tissue becomes firmer and loses its translucency and becomes increasing yellow-brown, the boundary between nucleus and annulus becomes more difficult to distinguish. In old age the hydra­ tion of the nucleus approaches that of the annulus. Contained within this matrix are cells, the chondrocytes, which are actively maintaining and repairing it. The mean cell density in the disc is very low so that cells occupy only about 1-5% of the tissue volume. Because of the low cellu­ larity, the mechanical properties of the disc depend chiefly on the constituents of the matrix. However, activity of the cells is vital for maintaining the integrity of the tissue.

The annulus fibrosus
To the naked eye, the annulus fibrosus appears to consist of a series of concentric layers surrounding the nucleus pul­ posus (see Fig. 5.1). This banded appearance results from the intricate arrangement of fibrous lamellae, which will be discussed in more detail later. The annulus is less hydrated than the nucleus, and changes with age in this structure are not so apparent.

Proteog Iyca ns
Proteoglycans (PGs) are some of the largest and most com­ plex molecular structures in mammalian tissue and consist of polysaccharide chains covalently bound to a central pro­ tein core. The number and type of polysaccharide chains and the organization, composition and size of the core pro­ tein are very variable. At present 30 different proteoglycans are known although not all occur in the disc. They fulfil a variety of different functions such as binding growth fac­ tors, regulating collagen fibril size and other properties, for example influencing transparency in the cornea (lozzo
& Murdoch 1996, Woods & Couchman 2001). However, in

The cartilage e nd-plate
In children this region acts as a growth plate until skeletal maturity is reached, when the outer ring of 2- 3


and fuses with the rim of the vertebral body. A plate of hya­ line cartilage, approximately 1-2 mm in thickness, remains abutting the central region of the disc throughout adult life. Fibres from the disc continue into the end-plate where they align horizontally. At the bony interface there is a region of calcified cartilage. The composition of the end-plate resem­ bles that of the disc, but with less water and a greater fibrous component (Roberts et aI 1989).

this chapter only the biophysical functions of the proteogly­ cans will be discussed. These relate principally to the response of the disc to mechanical load. As in all load-bearing cartilages, proteoglycans endow the matrix with a high osmotic pressure and a low hydraulic permeability and hence constitute the compression-resisting component of the disc.

Disc proteoglycans Aggrecan. The main proteoglycan found in the disc is the
large aggregating proteoglycan, aggrecan (JohD.stone
& Bayliss 1995). The protein core of aggrecan is able to

The matrix of the intervertebral disc is very similar in com­ position to that of articular cartilage. It consists of collagen

attach at one end to hyaluronan (HA), a long unbranched polysaccharide. Aggrecan, when undegraded, thus exists in

Chemistry of the intervertebral disc in relation to functional requirements


the tissue as very large aggregates consisting of a long chain of HA with many aggrecan molecules attached (see Fig. 5.2). The protein core of aggrecan has a second globular domain adjacent to the HA binding region followed by a long straight central domain. The glycosaminoglycan (GAG) chains of the polysaccharides chondroitin sulphate (CS) and keratan sulphate (KS) are covalently attached to the central domain of the protein core. Each intact aggrecan molecule has around 100 GAG chains attached to it (Iozzo
& Murdoch 1996).

substance, but a family of proteins of which there are approximately 21 members identified so far, with varying chemical composition and tertiary structure. There are at least 29 genes controlling the production of all these types of collagen. All members of the collagen family have three amino acid chains with at least some part of their molecule having these three chains coiled together forming a super helix. Chemical bonds, or cross-links, form both within the collagen molecule and between adjacent molecules. The physical and mechanical properties depend on the amino acid composition since this defines the tertiary structure and conformation of the protein and also its interactions with other collagen and matrix molecules. The collagen types can be classified into subgroups depending on their structure or organization. One example of such a classification is shown in Figure 5.3. Fibrillar col­ lagens (types I, II, III, V and XI), as their name suggests, form fibrils. The extensive helical structure of fibrillar colla­ gen and cross-links which form render these collagens very stable. They are highly insoluble, resistant to enzymatic breakdown, and are mechanically strong. Two of them, types I and II, are the most common collagen throughout the body, making up more than 80% of the total collagen. Type I collagen is found in tissues such as tendon, ligament and the outer region of intervertebral disc but is also the major collagen of skin and bone. Type II collagen is mainly found in tissues with a high degree of compressive loading, such as cartilage and the eardrum. Exactly how the mechanical properties of these different types relate to their chemistry remains unclear.

In situ in the disc, aggrecan molecules tend to be smaller than those from hyaline cartilages. Only about 30% of monomers found in the disc nucleus can form aggregates compared with about 80% in hip cartilage. Moreover, the aggregates from the nucleus are smaller, having a molecu­ lar weight of about 7 million compared to 100 million for aggregates found in bovine nasal septum. There are indi­ cations that the disc PGs are able to form aggregates when newly synthesized, but that they are degraded in the tissue and that their HA-binding region disappears (Johnstone & Bayliss 1995). The functional significance of differences in degree of aggregation and of PG size is not yet understood. It has been suggested that the aggregation helps to keep the PGs in the tissue since there is no evidence that they are held there by any form of binding.

Versican. Versican is a large proteoglycan of similar
domain structure to aggrecan and also is able to form aggregates. However, it has far fewer GAG chains attached to its protein core. Versican is found mainly in tissues such as ligaments and tendons. It is present in the annulus but not the nucleus of the disc (Hayes et al 2001, Melrose et a1 2001).

Small proteoglycans. The disc also contains several mem­
bers of the family of small leucine-rich proteoglycans. These consist of a short protein core with only one or two GAG chains attached to it. To date, decorin, biglycan, fibro­ modulin and lumican have been identified in the disc, all at higher concentrations in the annulus than in the nucleus apart from lumican. Decorin and fibromodulin bind to the outside of collagen fibrils and thus have roles in regulating collagen fibril diameter. Biglycan is known to bind growth factors Gohnstone et al 1993, Sztrolovics et al 1999).
200nm L.......J

Collagens I, II, III, V, XI N.
• •

C 1°

NC4 GAG chain

Collagens IX, XII, XIV



� �\ b . L\--


II -

Collagens IV 7S �.,--"". NC1

Collagens VI

Fixed-charge density (FeD)
One important property of the GAGs, especially in relation to the disc's mechanical function, is that they are charged. Both CS and KS contain anionic groups (S03- and COO-), which impart a net negative charge to the matrix. The FCD confers important properties on the disc since it controls the distribution of charged solutes and hence osmotic pressure, as discussed later.

•• •

• •

• • • • • •

• •

200 nm Collagens VII 100nm

beaded filament Collagens VIII, X NC 2

..-- NC1

. .. ..

Collagen is the main structural protein in the body, making up about 80% of the total body protein. It is not one single

basement membrane


Schematic showing the organization of the different

collagen types.



Collagen types in disc
There are 10 types of collagen identified to date in the disc. Types I and II are complementary, approximately 80% of the outer annulus being type I collagen, the amount decreasing centrally and vice versa for type II collagen. The other types present in the disc include fibrillar collagens V and XI, other minor collagens including types III and V I, which are mostly cell associated, the basement membrane collagen, type Iv, and the fibril-associated collagens with an interrupted triple helix (FACIT) collagen, types IX, XII and XIV. The FACIT collagens bind to the surface of type I and II collagen and may control their fibril diameter. Hence, some of these collagens, while only making up a small per­ centage of the total, may be a potent influence on the mechanical properties of the tissue.

Nucleus pulposus

Annulus fibrosus

Collagen organization in the disc
The organization of collagen fibrils in the disc is highly spe­ cialized. The three-dimensional collagen framework of the disc has been shown in scanning electron micrographs (Takeda 19 75). In the nucleus the collagen fibrils are much finer than in the annulus, mostly about 0.05 ).lm in diameter, and are arranged in a loose irregular meshwork. In the annulus, collagen is arranged in 15-25 concentric lamellae made of parallel bundles of fine fibrils 0.1-0.2 ).lm in diam­ eter. These lamellae are visible to the naked eye and vary from 100 to 500 ).lID in width, the outer lamellae being thick­ est. Their width varies with age and location, with the pos­ terior lamellae not as wide as those in the rest of the disc (Marchand & Ahmed 1990). The fibre bundles of each lamella run obliquely between the adjacent vertebral bodies and are firmly anchored to them or to the cartilaginous end­ plate. The resulting angle formed between the fibre bundles of the lamellae and the vertebral bodies varies between 40 and 70 degrees, the direction of the fibres alternating in the neighbouring lamellae. The arrangement of the collagen network in the disc has an important influence on how load is distributed. The angle between the fibre bundles of the adjacent lamellae is able to change since the lamellae are loosely interconnected. Even though collagen is only slightly extensible, the fact that the lamellae can move separately gives the structure itself considerable extensibility, especially in the vertical direction. The arrangement of the collagen network is shown schematically in Figure 5.4.




Schematic view of collagen organization of the disc

showing the direction of the collagen bundles in adjacent annulus lamellae and possible changes in the network on loading.

also found in disc tissue and may be involved in ageing and degenerative processes. The proportion of these compo­ nents changes with development and ageing and also varies with position (Nerlich et a1 1997, Hayes et aI 2001). The con­ centration of some proteins such as fibronectin increases markedly in degenerate discs (Oegema et al 2000).

The disc, like all other tissues, has to have the capability to remodel itself, which entails both breaking down or degrading the original matrix and synthesizing new com­ ponents. Proteinase enzymes, which can break down colla­ gen, proteoglycans and other matrix macromolecules have been identified in the disc (Sedowofia et aI 1982). Some pro­ teinases, such as cathepsins B and D, are more active at acid pH, while others, including matrix metalloproteinases and aggrecanase, are more active at neutral pH. Matrix metallo­ proteinases (MMPs), a family of enzymes each able to degrade specific macromolecules and together able to break down all matrix components, have been found in the disc. Only members of the subgroup collagenases (MMPs 1, 8 and 13) can disrupt the triple helix of collagen, but the degraded collagen fragments can be further broken down by other MMPs, the gelatinases (MMPs 2 and 9). The pro­ teoglycan core protein is broken down from both ends by many MMPs but particularly the stromelysins (MMPs 3, 10 and 11). However, the most common degradation site is near the HA-binding region, leaving a stub of aggrecan attached to HA together with non-aggregated aggrecan fragments. The activity of MMPs is complex. MM'ps are produced by the cells as inactive precursors which require activation before they can degrade matrix molecules. They also have naturally occurring inhibitors (tissue inhibitors of MMPs or TIMPs).

Other matrix constituents
In addition to collagen and PG the disc contains a consider­ able fraction of non-collagenous proteins. These include structural glycoproteins, such as elastin (Mikawa et a1 1986) and other less well-characterized constituents. Some may be associated with the cell as has been found in other cartilages, where they are strongly implicated in the interaction of the cell with the extracellular matrix. Amyloid, extravascular plasma proteins and endogenous proteinase inhibitors are

Chemistry of the intervertebral disc in relation to functional requirements


In degenerate discs :MMP concentrations increase (Crean et
al 1997), as shown in Figure 5.5A, and turnover, which is nor­ mally slow (> 2 years for PGs; > 10 years for collagen), is more rapid in degeneration (Antoniou et al 1996a). The increased proteoglycan degradation and :MMP activity seen in disc dis­ ease could result not only from an increased synthesis of :MMPs, but also from greater activation of those present or from decreased levels of TIMPs (Roberts et al 2000).

There are distinct differences between cells in the annulus and nucleus, possibly reflecting their different developmen­ tal origin. In the infant human disc, the cells of the nucleus are notochordal. These notochordal cells disappear by 4-10 years in humans and are replaced by rounded chondrocyte­ like cells surrounded by a capsule with long processes extending into the matrix. The cells of the inner annulus are also rounded, but those in the outer regions of the annulus are long and thin and extend along the collagen fibrils of the lamellae (Errington et al 1998). While disc cells have not yet been fully characterized, it is now clear that there are differ­ ent cell types, which remain distinct in culture and produce different sets of extracellular matrix proteins.

The disc has a low cellularity; the mean cell density of the adult human disc is about 5500 cells mm-3 (Maroudas et al 19 75). The cell density is not uniform throughout the tissue, being highest near the end-plate and at the periphery of the annulus, i.e. in the regions nearest the blood supply. The primary function of these cells is the manufacture and maintenance of the matrix.

Water containing dissolved solutes is the main constituent of the disc. It occupies 65-90% of the tissue volume, depending on age and region. Since the cell density is low, most of the water is extracellular. Some of it is associated with the collagen fibrils, the intrafibrillar fraction. This frac­ tion, which is about 1.0 g water/1.0 g dry collagen, may be considerable in old or degenerate discs, which have low PG concentrations. This intrafibrillar water is freely exchange­ able and is accessible to small solutes such as glucose, but large molecules, such as the PGs themselves, are excluded from this fraction. The effective concentration of PGs is then often underestimated (Maroudas 1990).

Nerves a nd blood vessels

Compared to other sites in the body, the intervertebral discs have few blood vessels or nerves in the healthy adult. In contrast, in utero and in young discs, there are blood ves­ sels in the annulus fibrosus and large vascular channels in


the thick cartilage end-plate. These diminish during devel­ opment. The cartilage end-plate decreases to become a thin layer, approximately 1.5 mm thick, of totally avascular hya­ line cartilage. Few, if any, vessels remain present in the outer annulus. The adult disc is almost avascular; it is the largest avascular structure in the body. However, with increasing age and/ or degeneration, the vascularity of the annulus increases again (Kauppila 1995) (Figure 5.58). Innervation in healthy and diseased discs follows a simi­ lar pattern to blood vessels. The normal adult disc has nerve fibres in the outer few millimetres of the annulus fibrosus. They originate from the sinuvertebral nerve and paraverte­ bral sy.mpathetic trunk, their distribution apparently being non-segmental, with reports of pain relief at the L4-S1 disc levels when anaesthetic is applied to the L 2 nerve root. The nerves lie predominantly parallel to the collagen bundles in the annulus and particularly between the annular lamellae, although sometimes nerve fibres can be seen to cross them.


Innervation is both autonomic and sensory, the former com­


A MMPs produced by cells in degenerate discs.

monly co-distributing with blood vessels in the disc, pre­ sumably when the nerve facilitates vascular control. The

B Blood vessels in annulus of degenerate discs.



sensory innervation may be involved in both nociception and proprioception, since it is known that mechanorecep­ tors occur in the outer 2 - lamellae of human discs and the 3 longitudinal ligaments (Roberts et al 1995). Golgi tendon organ receptors (GTO) appear to be the most common pro­ prioreceptors, although pacinian corpuscles and Ruffini endings have also been observed. The relative frequency of these mechanoreceptors may reflect the level and rate of change of stress seen in the disc, since GTOs are high thresh­ old, slow adapting receptors which only become active at the extremes of movement in other localities. The mechanoreceptors in the disc may be important in influenc­ ing the activity of spinal muscles. Electrical stimulation of the disc innervation elicits reaction in the lumbar multifidus and longissimus paraspinal muscles (Indahl et al 1997). In degenerate intervertebral discs removed from patients with back pain the innervation seen is more extensive than in normal controls, with both number and size of nerves increasing. Both thick myelinated nerve bundles (diameter 15-25 11m) and thin (diameter 0.25-2.5 11m) nerves are found in degenerate discs and nerves also penetrate more centrally in degenerate than in normal discs (Freemont et al 199 7).

The cause of disc herniation is not clear. It could arise from purely mechanical overload or because the matrix is weaker than normal. Poorly synthesized or structured matrix or matrix which has been partially degraded (e.g. via MMPs) could lead to a mechanically inferior matrix, which might predispose to rupture under normal loading conditions. The posterolateral disc bulges (protrusion) or ruptures either partially (extrusion) or totally (sequestration). Clinical symptoms result from pressure on the spinal nerves, depending at what level the prolapse occurs. L4-5 and L 5- S1 discs are most commonly involved affecting the sciatic nerve, resulting in sciatica. However, since more than 70% of individuals with disc pro lapses identifiable on MRI are asymptomatic (Boos et al 1995) there must be other factor(s) involved in addition to mechanical factors. Possible suggestions include sensitization of the nerve roots via inflammatory mediators which might be released from the herniated discs (Brisby et al 2000). The natural history of symptomatic herniated discs, if left unoperated, is often to resorb. They become vascular­ ized and have high levels of cytokines and proteases, such as matrix metalloproteinases, which can degrade the extra­ cellular matrix.

Disorders of the intervertebral disc include spinal deformi­ ties such as scoliosis (when the spine is bent laterally), kyphosis (when it is abnormally humped), disc herniation or disc degeneration. These will each be discussed in turn.

Disc degeneration
The disc shows degenerative changes relatively early in its life with about 10% of individuals aged 10-20 years having degenerate discs and with degeneration increasing steeply with ageing (Miller et al 1988). The causes are unknown but twin studies indicate that there is a strong genetic influence (Sambrook et al 1999). When twins discordant for the major risk factors such as heavy physical work or smoking were examined (i.e. one twin smoked but not the other) the influ­ ence of the environment was insignificant compared to that of the genes (Battie et al 1995). Disc degeneration appears to be a complex polygenetic disorder; associations between disc degeneration and polymorphisms of several genes such as aggrecan, collagen IX and vitamin D receptors have already been reported (Videman et al 1998, Annunen et al 1999, Kawaguchi et al 1999 , Paassilta et al 2001). Disc degeneration is thought to lead to back pain, either directly or indirectly. One of the main compositional changes in degenerate discs is loss of proteoglycan and hence water and actual disc height. The associated loss of stiffness leads to increased disc bulging on loading. These changes can in turn alter the loading on adjacent spinal structures such as muscles, ligaments, facet capsules and bone, all of which are highly innervated. It can contribute to the symptoms associ­ ated with spinal stenosis. In this condition, which occurs later in life, the spinal canal becomes narrowed due to increased ossification of the vertebral bodies bordering on the spinal canal and/ or the thickening and sometimes calci­ fication of ligamentum flavum. Slight bulging of the disc in such a narrowed canal will therefore cause symptoms.

Spinal deformities
In both scoliosis and kyphosis the intervertebral discs are wedged, as can be the vertebral bodies. There are many congenital, neuromuscular or other disorders which result in scoliosis, but in the majority of affected individuals there is no known cause (idiopathic). As far as is known the com­ positional and structural changes in the scoliotic disc are the same whatever the initiating factor. Several properties such as the type of collagen, its cross­ linking pattern and turnover and glycosaminoglycan con­ tents have all been shown to vary across the disc from the concave to the convex side of the curve (Crean et al 1997, Duance et al 1998). The scoliotic disc also shows marked calcification, particularly in the end-plate (Roberts et al 199 3). More recently, techniques used to study nutrient pathways have demonstrated that these are affected in sco­ liosis with those at the apex of the curve being most com­ promised (Urban et al 200 la). This may be directly responsible for the diminished cell numbers seen in these discs (Urban et al 2001b).

Disc herniation
Disc herniation or 'slipped' or prolapsed discs are the most common disc disorder leading to spinal surgery, with 12 000 operations for it being carried out annually in England.

Chemistry of the intervertebral disc in relation to functional requirements


In addition to the loss of proteoglycan and water, other biochemical changes which can occur in degenerate discs include·a shift in the types of collagen produced, altered collagen cross-linking and increased proteinase content. Morphologically, the disc becomes cracked and fissured, perhaps due to loss of glycosaminoglycan and water, with more irregular, disrupted collagen organization and annu­ lar lamellae. In addition degenerate discs have more blood vessels and innervation than 'normal' discs (Freemont et al 1997). The degeneration-induced changes in disc composi­ tion and structure can be visualized to some extent by MRI (Thompson et al 1990, Antoniou et aI1998).

erable in comparison with that of non-weight-bearing tis­ sues. It should be noted that the osmotic pressure rises steeply with an increase in FCD or PG content arising, for instance, as the result of fluid expression under load.

Swelling pressure of the disc
The disc contains PGs at concentrations which lead to an osmotic pressure, 1t, of several atmospheres. PGs at such concentrations in contact with saline solutions would tend to imbibe water and hence cause the tissue to swell. In vivo this tendency to swell is opposed by: (i) the combined effects of body weight and muscle tension, Pa; and (ii) the net restraining force of the collagen network of the disc, Pc. At equilibrium these effective pressures are balanced. Thus we can say: Pa = (1t - Pc) Ps (equation 1)

The major functions of the discs are mechanical as the discs serve both to transmit load and to act as joints. These mechanical functions are directly related to the concentra­ tion and arrangement of the two major structural compo­ nents of the tissue, collagen and Pc.


where Ps is called the net swelling pressure of the tissue. Ps varies with the composition and hence with region of the disc and will be affected by factors such as age and degen­ eration which affect tissue organization and composition and also with load-induced changes in hydration, since both 1t and Pc vary with hydration.

A structural model
The functions of the macromolecular constituents of the matrix are demonstrated in a physical model (Broom & Marra 1985). This model consists of a network of string enclosing balloons, which inflate the network and prevent it from collapsing. The resulting structure is able to support compressive loads though neither the string nor the bal­ loons could do so alone, as the string would collapse and the balloons would fly apart. In the disc, collagen forms a structural framework which, like the string network of the model, is strong in tension but collapses under a compres­ sive load if unsupported. PGs are held in the matrix by the collagen network as the balloons are held in the string net­ work PGs imbibe water, as discussed later, and inflate the collagen network, as the balloons inflate the string network; in so doing, they enable the tissue to support compressive loads without collapsing.

Swelling pressure and tissue fluid content
Fluid content depends on the applied load, on the loading history and on tissue composition. If the load on the disc at equilibrium is increased, the equilibrium is disturbed by the increase in Pa since now Pa>Ps; fluid is thus expressed to restore equilibrium. As fluid is expressed the PG concentration, and hence the osmotic pressure, increases but the volume of the tissue, and hence the collagen network tension, decreases. The net effect is an increase in the swelling pressure Ps (equation 1). If the load is maintained, fluid will continue to be expressed from the tissue until the swelling pressure increases suffi­ ciently to balance the applied pressure. Conversely, if the load on the tissue is reduced Pa< Ps and the disc swells. During swelling, the PGs are diluted and their osmotic pressure, 1t, decreases; as the volume increases, the tension in the collagen network also increases and Pc rises. The swelling pressure is consequently reduced. If the load on the disc is completely removed, for instance when the disc is placed in saline solution in vitro, swelling will finally cease when the reduced osmotic pres­ sure is balanced by the increased collagen tension. Figure 5.6 shows the increase in volume of a bovine tail disc after incubation in saline for 5 hours. Swelling was con­ siderable, especially in the nucleus where the collagen net­ work is weakest; here the fluid content of similarly swollen discs may increase 200- 300% (Urban & Maroudas 1981). Swelling in the annulus was directional and dictated by the collagen organization. It should be noted that following disc herniation, sequestrated disc fragments are in effect unloaded fragments of disc, which can swell considerably.

The biophysical function of proteoglycans and collage n

The osmotic pressure of proteoglycan solutions
The high osmotic pressure of PG solutions arises mainly from the polyelectrolyte nature of the PGs. Because of the fixed negative charges on the GAGs, the total number of ions in the disc is always greater than in the plasma and the excess number of ions in the disc leads to the high osmotic pressure in the tissue. PG size or degree of aggregation has little influence on osmotic pressure compared to charge density (Comper & Preston 1974). In the nucleus of the rest­ ing adult disc the FCD lies between 0. and 0.4 mEq ml-1 2 depending on age, giving an osmotic pressure of 0.1-0.3 MPa (1-3 atmospheres). The FCD of the annulus, and hence its osmotic pressure, is somewhat lower, but is still consid-





from the normal disc matrix on account of their size. Even glucose (mol wt <200) is sterically excluded from about 10% of the pores in a normal disc (Urban et al 1979). In disc degeneration, as PG concentration falls, pore size increases

annulus --I!o(a)



and some of the pores become accessible to these large mol­ ecules. This process has not been investigated in the disc, but loss of PGs allows large molecules such as growth fac­ tor complexes to penetrate into osteoarthritic articular car­ tilage, possibly influencing the development and progression of arthritis (Schneiderman et al 1995). Proteoglycans also govern the movement of solutes through the tissue. Solutes can move by diffusion under concentration gradients or through convective flow, i.e. with fluid moving in and out of the disc in response to changes in mechanical load. In both cases, their movement



Swelling of the unloaded disc. A An intact bovine tail

disc after removal of adjacent tissues. B the same disc after immersing for

5 hours

in saline at

4T. The

nucleus pulposus in

particular has swollen considerably; swelling has increased the height of the annulus fibrosus more than the width.

Hydraulic permeability
In the schematic view of the disc matrix shown in Figure 5.2 the matrix can be seen to consist of PGs densely packed between collagen fibrils. The collagen fibrils are spaced at 20- 60 nanometres whereas, because of their close packing, the distance between the GAG chains is only 2 - 4 nano­ metres (Byers et al 1983). The fine pore structure of the matrix is thus determined by the PG concentration rather than by the collagen network; the higher the PG concentra­ tion, the more closely packed are the GAGs and the smaller the effective 'pores' formed by the entangled GAG chains. A change in the water content of the disc alters PG concen­ tration and thus pore size; if the tissue swells, the PG con­ centration is diluted and the effective pore size increases. Conversely, if the disc loses fluid, the pore size decreases as the same number of PG chains pack into a smaller volume of tissue. Pore size relates directly to hydraulic permeability and thus the rate at which fluid moves in and out of the disc; small pores impart a low hydraulic permeability and thus restrict fluid flow.

is restricted by the GAG network. Whether a solute moves by diffusion or whether transport is aided by convection depends on solute size. Small solutes such as glucose, oxy­ gen and lactate, which can diffuse rapidly, move almost entirely by diffusion; larger solutes such as enzymes or pro­ teases, which diffuse more slowly, are affected by fluid movement (Thornton et al 1987, O'Hara et al 1990).

The disc's behaviour under load depends both on the load and on the organization of its extracellular matrix.

Loads on the lumbar spine
In vivo the disc is always under load as a result of the com­ bined effects of body weight and muscle activity. Loads applied to the disc lead to a rise in pressure. Nachemson (19 60) found that even in a relaxed supine subject the pres­ sure on the lower lumbar discs was 0.1-0.2 MPa, whereas in unsupported sitting it rose to about 0.6-0.7 MPa. Peak pres­ sures during strenuous activity may rise considerably above these values. Recent in vivo measurements have monitored changes in intradiscal pressure over 24 hours and confirmed these early results for the most part (Wilke et al 1999), with the pressure lower in degenerate than in normal discs (Sato et al 1999). Because of the relationship between disc pressure and posture, the intradiscal pressure tends to follow a cyclic pattern: it is at its lowest during sleep, and then increases 5- to 6-fold during the day's activ­ ities.

Solute partitions and transport
Proteoglycans, because they are charged and divide the extracellular spaces into small pores, control the concentra­ tion of dissolved solutes in the disc. The imbalance of charge in the matrix leads to the concentration of positive ions, such as calcium and sodium, being higher in the disc than in the external plasma (Maroudas 1980). The distribution of ions is important because it governs the osmotic pressure of the disc as discussed. Charge also affects the concentration of other molecules which enter the disc, such as antibiotics; positively charged antibiotics such as aminoglycosides will reach higher concentrations in the disc than negatively charged antibiotics such as penicillin (Thomas et al 1995). The concentration of large uncharged molecules in the disc is governed by the pore size distribution and this also depends on PG concentration. Many of the larger serum proteins have diameters which are greater than the 2 - 4 nanometre diameter o f the 'pores' formed between the interdigitating GAG chains and are thus virtually excluded

The deformation of the disc under load
Disc height alters with changes in load by two mechanisms. When the load on the disc is increased the disc deforms ini­ tially through a rearrangement of the collagen network (see Fig. 5.4). The extent of this deformation varies from disc to disc, but the factors which govern this are not understood. No consistent pattern with age, sex or degree of oegenera-

Chemistry of the intervertebral disc in relation to functional requirements


tion has been found (Shirazi-Adl et al 1984). For changes in load of short duration this deformation is virtually constant volume.· However, if the load is maintained the disc loses height or creeps, and a large part of the creep deformation results from fluid loss (Keller & Nathan 1999). Thus, for each applied load, the extent of the initial deformation will depend largely on the structure and integrity of the colla­ gen network (Iatridis et al 1999). In contrast, the rate and magnitude of the creep deformation is related more to the PG content. Degenerate discs of low PG content will be much less able to retain fluid in the face of applied pressure than normal discs (Fig. 5.7A). The amount of water exchanged between disc and sur­ roundings with changes in load is quite considerable. In vivo MRI studies which imaged discs of normal office worker volunteers immediately after rising and then after a day's work estimated that, on average, discs lose around 25% of their fluid content during the day and that this fluid is reimbibed at night during rest (Boos et al 1993). A schematic view of the diurnal fluid cycle and its rela­ tionship to swelling pressure is shown in Figure S.7B. Initially the disc is assumed to be at equilibrium at point (1) (after a night's rest - 7 a.m.). The pressure on the disc is suddenly increased to (2) (on rising, for example). The disc is no longer in equilibrium. Thus there is a driving force expressing fluid from the tissue which depends on the dif­ ference between (2) and the swelling pressure curve. While the pressure on the disc is maintained, fluid is expressed from the tissue. The rate of fluid loss depends partly on the driving force (it is fastest when the driving force is greatest) and partly on the hydraulic permeability of the tissue at point (2). Fluid loss is thus fastest initially. As fluid is expressed the PG concentration rises and the hydraulic per­ meability falls; also, the driving force diminishes as the equilibrium curve is approached. All these mechanisms help to limit the amount of fluid lost from the disc. When the pressure on the disc is released (points 3 to 4 - on retir­ ing to rest at night, for example), the disc now lies under the swelling pressure curve and thus will have a tendency to imbibe fluid in order to dilute the PGs. During swelling (during rest at night - points 4 to 1) the hydraulic perme­ ability increases; thus, the rate of fluid flow does not decrease drastically as equilibrium is approached. Swelling thus tends to be faster than fluid loss and hence the disc is able to replace the fluid lost in 16 hours of activity with 8 hours of rest. Changes in height during the day are thought in part to arise from loss of fluid from the disc. Eklund & Corlett (1984) measured 6 mm height loss on average during the day, and found that the rate of shrinkage depended on the load on the spine. MRI measurements have correlated this loss in height with fluid loss (Paajanen et al 1994). In con­ trast, Thornton et al (1987) found that the Skylab astro­ nauts, who were weightless for 85 days, grew about 5 cm in height, in part probably through swelling of their discs under low external loads.
06 . 0.8 20 60 100 Time, mins 140 180 220
Q) " >



� :5 70 0 #:E C> ;:

'" Cl. 6
i!? i!? Cl.
" '" '"

. 04

0.2 Swelling pressure curve

2 4 Hydration (gHPlg dry tissue)


Disc fluid content depends on disc composition and on

load. A Rates of fluid expression from disc slices in vitro in relation to disc proteoglycan content (open circle: high initial PG content; solid circles: low initial PG content). B Schematic of the diurnal fluid cycle in relation to the disc swelling pressure curve. at rest;


7 a.m.:

(2) 7.05 a.m.: pressure increase retiring; (4) pressure drop on retiring.

on rising;


11 p.m.: before

Although the disc has a low cell density, the continuing activity of the cells is vital to the health of the disc since the cells are responsible for turning over and renewing matrix constituents.



Energy metabolism
Disc cells produce the energy necessary for their survival and function mainly from glucose, which is broken down by glycolysis, even in the presence of oxygen, to produce lactic acid. Through this pathway one molecule of glucose produces two molecules of lactic acid and two molecules of ATP, the cellular energy source. ATP can be produced more efficiently by oxidative phosphorylation since this pathway, which requires oxygen as well as glucose, produces up to 36 molecules of ATP from one molecule of glucose. However, it seems that only around 15% of the disc's ATP is produced by oxidative phosphorylation (Holm et al 1981) and that disc cells do not need oxygen to survive (Homer & Urban 2001). Without oxygen, however, cell functional activity falls steeply.

Waste products produced by the cells are removed from the tissue by the same blood vessels. The route through the end­ plate may disappear if a calcified layer forms, as happens in scoliosis (Roberts et al 1996) or disc degeneration (Nachemson et al 1970); the nucleus cells may then be at risk. Because of its size and avascularity, steep gradients in the concentration of nutrients exist in the disc. Holm et al (1981) have shown that while the outer annulus is in equi­ librium with the blood oxygen, in the disc interior oxygen concentrations are very low and lactic acid concentrations high so that the deep regions are acidic. The disc thus appears to be in a precarious metabolic state with any loss of nutrient supply leading to cell death and disc degenera­ tion. Indeed, factors affecting the blood supply to the verte­ bral body, such as atherosclerosis (Kauppila 1997, Kauppila et al 1997), sickle cell anaemia, caisson disease and Gaucher's disease Gones 1997), all appear to lead to a sig­ nificant increase in disc degeneration. In experimental tests, factors which affect the microcir­ culation, such as vibration or smoking, were seen to lead to a rapid fall in oxygen concentrations in the disc nucleus and a rise in lactate concentration (Holm & Nachemson 1988; see Figure 5.8B). Long-term exercise - or lack of it - appears to have a permanent effect on movement of nutrients into

Disc nutrition
The disc is avascular. Nutrients such as oxygen and glucose, essential for cell survival and activity, reach the cells by dif­ fusion through the matrix of the disc from the blood vessels in contact with the annulus periphery and with the cartilagi­ nous end-plate, as shown schematically in Figure 5.8A.

Figure S.S

A Schematic view of the nutritional pathways into the

disc and of the blood vessels at the vertebral body-disc interface. Adapted from Holm et al


B Effect of exposure to smoke on the

time course of (i) oxygen tensions and



lactate concentrations

J/ JJJl J/

measured in vivo in the centre of the nucleus pulposus (open squares: exposed to smoke; closed circles: no smoke control). Adapted from Holm & Nachemson

1 988.


1 00 80
.. "�


c 0 'in c .l!l c Q) 0> �

60 40 20

'::" '0 E -= c o


� �

0 20 40 60 80 1 00

g 8




2-r------,--, 60 ioo 40 o 20 80
Time (minutes)

Time (minutes) --.- No smoke -0- Exposed to smoke

Chemistry of the intervertebral disc in relation to functional requirements


the disc and thus on their concentration in the tissue. The mechanism is not known but it has been suggested that exercise. affects the external capillary bed at the disc-bone interface (Holm & Nachemson 1983). Holm & Nachemson (1982) examined dogs' discs which had been fused. After 3 months a fall in cellular activity could be observed, PGs were lost and the fluid content of the discs fell. The reverse occurred in dogs which were vigorously trained over sev­ eral months; this resulted in increased cellular activity and PG content. It is, however, not clear whether the effect of load or exerdse was entirely due to change in nutrient sup­ ply; as will be discussed below, cell metabolism is also very sensitive to mechanical stress.

Animal models do indeed suggest that PG replacement is possible. When dog discs were treated with chymopapain and PGs were lost from the disc and end-plate, it was observed that the undamaged disc cells were able to syn­ thesize PGs, and expansion of the disc was observed after several months (Garvin & Jennings 1973, Oegema et al 1983). There is no evidence, however, that PGs can be replaced after chymopapain treatment in humans (L eivseth et al 1999), probably because in damaged human discs many disc cells are lost before treatment is started.

Extracellular influences on disc cell metabolis m
Over the last 10 years development of methods for study­ ing matrix metabolism (Bayliss et al 1986, Maldonado

In vivo measurements
Relatively little is known about matrix synthesis and turnover in the disc. In vivo studies in animals using radioactive labelling demonstrated that PGs are synthe­ sized in vivo in both adult and young animals. Turnover time (i.e. the average time to replace all PGs) was only a few weeks in 6-week-old guinea pigs, but was over 2 years in adult dogs (L ohmander et al 19 73, Urban et aI 19 79). Recently, new techniques have been used to examine turnover in discs obtained from human surgical and autopsy material. These rely on examining tissue-specific markers of breakdown or synthesis using newly developed antibodies. Breakdown can be measured by antibodies to neo-epitopes, produced when molecules such as aggrecan or collagen are cleaved by proteinases (Hughes et aI1998). Also, tissues can be examined for molecules produced only during synthesis. For instance, after the collagen molecule is exported from the cell but before it can be assembled into the matrix, a protein domain, the propeptide, is removed enzymically. These propeptides are relatively small and dif­ fuse from the tissue within days. Their presence in the tis­ sue thus indicates that there is active collagen synthesis. Antoniou et al (1996a, 1996b) have used these markers to examine discs over a range of degenerative grades and ages. Although results varied from region to region of the disc, in general they were able to identify three matrix turnover phases. Phase I (growth) was characterized by both active synthesis and active degradation of matrix mol­ ecules. Phase II (maturation and ageing) was distinguished by a progressive drop in both synthetic activity and denat­ uration of type II collagen. Phase III (degeneration) showed a fall in aggrecan and collagen II synthesis but an increase in collagen II degradation and in collagen I synthesis. Tracer measurements on human discs removed at surgery are in agreement, finding that PG synthesis varies across the disc and was low in degenerate discs Gohnstone
& Bayliss 1995).

& Oegema 1992 ) has increased understanding of the

factors influencing cellular activity. Disc cells appear to make a variety of matrix macromolecules but rates of synthesis vary depending on the cell origin; nucleus cells, for instance, produce aggrecan at much higher rates than outer annulus cells. Matrix synthesis also varies with age; rates of biosynthesis are fastest in cells taken from immature discs. Using these culture methods, it has become apparent that disc cells respond to a variety of extracellular stimuli and that the matrix produced depends not only on cell ori­ gin, but that the extracellular environment also has a pow­ erful influence on cell metabolism.

Growth factors a nd cyto kines
Disc cells respond to growth factors, such as IGF-1, which are responsible for stimulating matrix production (Thomp­ son et al 1991, Osada et al 1996). They also respond to cytokines such as IL-1 and TNF-a, which both stimulate activity of MMPs and other agents involved in matrix break­ down and repress synthesis of matrix macromolecules. The concentration of cytokines increases in herniated tissue (Kang et al 1996), possibly because inflammatory cells invade and populate the protruding disc. These cytokines may have a positive role to play in stimulating resorption of the protrusion; however, they may also set off a degenera­ tive cascade in the disc itself and possibly also stimulate pain in the nerve fibres in the outer regions of the disc.

Nutrient levels
Several studies have now shown that if nutrient supply to the disc is impeded, concentrations of oxygen and glucose in the centre of the disc fall and concentrations of lactic acid rise so that the disc becomes acidic (Diamant et aI19 68). In acidic pH or low oxygen, even if the cells survive, the amount of PG produced falls significantly (Ishihara
& Urban 1999). Even though PG turnover is slow, a

Since disc cells are active throughout life, potentially they might be able to repair the disc after injury or damage.

decrease in rate of production will eventually lead to a fall in PG concentration in the tissue with consequent changes



in disc biomechanics; loss of PG appears to be one of the first signs of disc degeneration.

Mec hanical stress
The disc is under constantly varying mechanical forces. With every movement or change in posture, the load on the disc alters as can be seen in recent continuous in vivo pressure measurements (Wilke et al 1999). Cells of most tis­ sues are very responsive to mechanical forces and recent work has shown that this is also true for disc cells.

In vivo responses to load
Most of the information in vivo comes from experimental studies where animals or joints have been subjected to abnormal mechanical loads for days to months. Little is known about the effects of exercise as such, though heavy exercise (40 km running per day) appeared to stimulate matrix synthesis marginally in dog discs (Puus�arvi et al 1993). However, abnormal loads appear to have detrimental effects. Spinal fusion, for instance, appears to lead to degen­ erative changes in adjacent discs. Degenerative changes and cell death have also been seen after discs have been subjected to high continuous compressive loading (Higuchi et al 1983, Lotz et al 1998). Long-term wedging can also produce disc abnormalities (Pazzaglia et al 1997). These studies all indi­ cate that degenerative changes can be induced by abnormal forces on an otherwise healthy disc and that these changes result from alterations in cellular activity rather than from matrix damage as such. While some of the effects of load in vivo might arise from alterations in the blood supply to the disc, in vitro tests have shown that disc cells themselves respond to load-induced changes in their environment. respond to each of these signals via different pathways. Nucleus cells, for instance, are very responsive to hydro­ static pressure; pressure in the low physiological range (0.3 MPa) stimulates PG synthesis significantly, whereas high pressure (3 MPa) inhibits PG synthesis but stimulates pro­ duction of MMPs. The effect of hydrostatic pressure appears to be mediated in part by nitric oxide (Liu et al 2001). Disc cells are also very sensitive to changes in hydra­ tion, with synthesis rates showing a bimodal response to load (Fig. 5.10); rates fall if fluid is expressed

Deformation changes the organization of the cytoskeleton

Fluid expression increases concentration of matrix PGs and other macromolecules around the cell



Schematic showing the effects of load on the

environment of the cell. On loading the disc matrix and cell deform, hydrostatic pressure rises, fluid is expressed, thus changing the composition of the matrix arou.nd the cell.

if the disc

In vitro studies
While in vivo studies have demonstrated an overall response of the disc cells to mechanical signals, under­ standing of the precise mechanical signals which stimulate the cells can only be obtained from in vitro experiments where specific responses to controlled mechanical signals can be investigated. Few studies on disc cells have so far been reported. However, results have shown that disc cells are very sensitive to mechanical stress and responses depend both on the cell type and on the precise nature of the mechanical signal. The type of mechanical signals seen by the cell depends on how the disc is loaded. When the matrix is loaded, hydrostatic pressure rises, the cell and matrix deform and fluid is expressed. Fluid moves along the cell boundary and, as a consequence of fluid expression, the extracellular concentration of macromolecules increases. The change in pressure or extent of fluid loss depends on the magnitude and duration of the load and on the disc composition. The signals seen by the cells on each change of load are thus very complex, as indicated in Figure 5.9. In vitro tests have shown that disc cells, as those of other cartilages, are sensitive to the magnitude of the load and
F i gure
--.- Nucleus
Q) -

18 16 14 "§ � 12

4 2 O -r------,--, o 60 80 40 20
Load (kg) --D-- OA

5.1 0

Effect of compressive load on proteoglycan synthesis

by cells of the bovine nucleus pulposus and annulus fibrosus. The whole disc was incubated at synthesis measured over

3TC in vitro

under load and rates of


hours (solid circles: nucleus; Open

squares: outer annU l US). Adapted from Ohshima et al 1 995.

Chemistry of the intervertebral disc in relation to functional requirements


swells (Ohshima et al 1995). Here the signal appears medi­ ated by the change in cell volume. Responses to fluid move­ ment and to stretch, however, appear to be regulated by cell- matrix interactions. The complexity is increased because the response varies with cell type. Annulus and nucleus cells have been shown to respond differently to the same mechanical signal in sev­ eral studies. For example, only nucleus and inner annulus cells are affected by a rise in hydrostatic pressure; outer annulus cells show no response to even high levels of pres­ sure (Ishihara et al 1996). Figure 5.10 shows that annulus cells produce less PG than nucleus cells and are less influ­ enced by compressive load. These in vitro tests demonstrate the sensitivity of disc cells to different mechanical signals. Load-induced changes in the extracellular environment of the cell alter production of matrix macromolecules and of proteases and hence can affect the overall composition of the disc in the long term. However, while in vitro tests are able to examine cellular responses to simple signals such as controlled stretch or fluid movement or pressure rise, in vivo the cell will be exposed to simultaneous changes in all these signals. Each signal will vary in duration and magnitude depending on the loading regime and nature of the matrix. The overall response of the cell to load thus depends on how it inte-

grates these different signals to produce extracellular matrix. At present we understand little of this process. Thus, although it is apparent that mechanical loading can affect the disc matrix in the long term, at present we are far from being able to predict the net response of the disc cells to any mechanical intervention.

In order to function adequately, the disc must retain a well­ ordered extracellular matrix throughout life. Disc cells make and maintain this matrix; any loss of cellular function will eventually lead to loss of matrix components and disc degeneration. At present, we understand little of the behav­ iour of these cells. We need to understand more about their behaviour in health and disease in order to preserve their activity, prevent disc degeneration and even possibly pro­ mote disc repair.

nucleus pulposus cartilage end-plate proteoglycans collage n protei nases disc nutrition

We thank the Arthritis Research Campaign (U0511) and the EU Consortium EURODISC(QLK6-CT-2002-02582) for support.

Annunen S, Paassilta P, Lohiniva J et al 1999 An allele of COL9A2 associated with intervertebral disc disease. Science 285: 409-412 Antoniou J, Steffen T, Nelson F et al 1996a The human lumbar intervertebral disc: evidence for changes in the biosynthesis and denaturation of the extracellular matrix with growth, maturation, ageing, and degeneration. Journal of Clinical Investigation 98: 996-1003 Antoniou J, Goudsouzian N M, Heathfield T F et al 1996b The human lumbar endplate: evidence of changes in biosynthesis and denaturation of the extracellular matrix with growth, maturation, ageing, and degeneration. Spine 21: 1153-1161 Antoniou J, Pike G B, Steffen T, Bararnki H, Poole A R, Aebi M, Alini M 1998 Quantitative magnetic resonance imaging in the assessment of degenerative disc disease. Magnetic Resonance in Medicine 40: 900-907 Battie M C, Videman T, Gibbons L E, Fisher L D, Manninen H, Gill K 1995 1995 Volvo Award in clinical sciences. Determinants of lumbar disc degeneration: a study relating lifetime exposures and magnetic resonance imaging findings in identical twins. Spine 20: 2601-2612 Bayliss M T, Urban J P G, Johnstone B, Holm S 1986 In vitro method for measuring synthesis rates in the intervertebral disc. Journal of Orthopaedic Research 4: 10-17 Boos N, Wallin A, Gbedegbegnon T, Aebi M, Boesch C 1993 Quantitative MR imaging of lumbar intervertebral disks and vertebral bodies: influence of diurnal water content variations. Radiology 188: 351-354 Boos N, Rieder R, Schade V, Spratt K F, Semmer N, Aebi M 1995 Volvo Award in clinical sciences. The diagnostic accuracy of magnetic resonance imaging, work perception, and psychosocial factors in identifying symptomatic disc herniations. Spine 20: 2613-2625 Brisby H, Byrod G, Olmarker K, Miller V M, Aoki Y, Rydevik B 2000 Nitric oxide as a mediator of nucleus pulposus-induced effects on spinal nerve roots. Journal of Orthopaedic Research 18: 815-820 Broom N D, Marra D L 1985 New structural concepts of articular cartilage demonstrated with a phYSical model. Connective Tissue Research 14: 1--8 Buckwalter J A 1995 Ageing and degeneration of the human intervertebral disc. Spine 20: 1307-1314 Byers P D, Bayliss M T, Maroudas A, Urban J, Weightman B 1983 Hypothesizing about joints. In: Maroudas A, Holbrow E J (eds) Studies in joint disease 2. Pitman, London, pp 241-276 Comper W D, Preston B N 1974 Model connective tissue systems: a study of polyion-mobile ion and of excluded-volume interactions of proteoglycans. Biochemical Journal 143: 1-9 Coventry M B, Ghormley R K, Kernohan J W 1945 The intervertebral disc: its microscopic anatomy and pathology; changes in the intervertebral disc concomitant with age. Journal of Bone and Joint Surgery 27: 233-247 Crean J K, Roberts S, Jaffray D C, Eisenstein S M, Duance V C 1997 Matrix metalloproteinases in the human intervertebral disc: role in disc degeneration and scoliosis. Spine 22: 2877-2884




Diamant B, Karlsson J, Nachemson A 1968 Correlation between lactate levels and pH in discs of patients with lumbar rhizopathies. Experientia 24: 1195-1196 Duance V C, Crean J K, Sims T J et al 1998 Changes in collagen cross­ linking in degenerative disc disease and scoliosis. Spine 23: 2545-2551 Eklund

Kang J 0, Georgescu H I, McIntyre-Larkin L, Stefanovic-Racic M, Donaldson W F, Evans C H 1996 Herniated lumbar intervertebral discs spontaneously produced matrix metalloproteinases, nitric oxide, interleukin-6, and prostaglandin E2. Spine 21: 271-277 Kauppila L 1 1995 Ingrowth of blood vessels in disc degeneration: angiographic and histological studies of cadaveric spines. Journal of Bone and Joint Surgery (American volume) 77: 26-31 Kauppila L I 1997 Prevalence of stenotic changes in arteries supplying the lumbar spine: a post-mortem angiographic study on 140 subjects. Annals of the Rheumatic Diseases 56: 591-595 Kauppila L I, McAlindon T, Evans S, Wilson P W, Kiel D, Felson 0 T 1997 Disc degeneration/back pain and calcification of the abdominal aorta: a 25-year follow-up study in Framingham. Spine 22: 1642-1647 Kawaguchi Y, Osada R, Kanamori M et al 1999 Association between an aggrecan gene polymorphism and lumbar disc degeneration. Spine 24: 2456-2460. Keller T S, Nathan M 1999 Height change caused by creep in intervertebral discs: a sagittal plane model. Journal of Spinal Disorders 12: 313-324 Leivseth G, Salvesen R, Hemminghytt S, Brinckmann P, Frobin W 1999 Do human lumbar discs reconstitute after chemonucleolysis? A 7-year follow-up study. Spine 24: 342-347 Liu G Z, Ishihara H, Osada R, Kimura T, Tsuji H 2001 Nitric oxide mediates the change of proteoglycan synthesis in the human lumbar intervertebral disc in response to hydrostatic pressure. Spine 26: 134-141 Lohmander S, Antonopolous C, Friberg U 1973 Chemical and metabolic heterogeneity of chondroitin sulfate and keratan sulfate in guinea pig cartilage and nucleus pulposus. Biochimica et Biophysica Acta 304: 430-448 Lotz J C, Colliou 0 K, Chin J R, Duncan N A, Liebenberg E 1998 1998 Volvo Award winner in biomechanical studies. Compression­ induced degeneration of the intervertebral disc: an in vivo mouse model and finite-element study. Spine 23: 2493-2506 Maldonado B A, Oegema-TR J 1992 Initial characterization of the metabolism of intervertebral disc cells encapsulated in microspheres. Journal of Orthopaedic Research 10: 677-690 Marchand F, Ahmed A M 1990 Investigation of the laminate structure of lumbar disc annulus fibrosus. Spine 15: 402-410 Maroudas A 1980 Physical chemistry of articular cartilage and the intervertebral disc. In: Sokoloff L (ed) The joints and synovial fluid II. Academic Press, New York, pp 240-293 Maroudas A 1990 Different ways of expressing concentration of cartilage constituents with special reference to the tissue's organization and functional properties. In: Maroudas A, Keuttner K (eds) Methods in cartilage research. Academic Press, London, pp 219 Maroudas A, Stockwell R A, Nachemson A, Urban J 1975 Factors involved in the nutrition of the human lumbar intervertebral disc: cellularity and diffusion of glucose in vitro. Journal of Anatomy 120: 113-130 Melrose J, Ghosh P, Taylor T K 2001 A comparative analysis of the differential spatial and temporal distributions of the large (aggrecan, versican) and small (decorin, biglycan, fibromodulin) proteoglycans of the intervertebral disc. Journal of An.atomy 198: 3-15 Mikawa Y, Hamagami H, Shikata J, Yamamuro T 1986 Elastin in the human intervertebral disk: a histolOgical and biochemical study comparing it with elastin in the human yellow ligament. Archives of Orthopaedic and Trauma Surgery 105: 343-349 Miller J, Schmatz C, Schultz A 1988 Lumbar disc degeneration: correlation with age, sex, and spine level in 600 autopsy specimens. Spine 13: 1 73-178 Nachemson A 1960 Lumbar intradiscal pressure. Acta Orthopaedica Scandinavica 43 (Suppl.): 1-104 Nachemson A, Lewin T, Maroudas A, Freeman M A F 1970 In vitro diffusion of dye through the end-plates and annulus fibrosus of

J A, Corlett E N

1984 Shrinkage as a measure of the effect of

load on the spine. Spine 9: 189-194 Errington R J, Puus�arvi K, White I F, Roberts S, Urban J G 1998 Characterisation of cytoplasm-filled processes in cells of the intervertebral disc. Journal of Anatomy 192: 369-378 Freemont A J, Peacock T E, Goupille P, Hoyland J A, O'Brien J, Jayson M-I V 1997 Nerve ingrowth into diseased intervertebral disc in chronic back pain. Lancet 350: 178-181 Garvin P Jennings R B 1973 Long term effects of chymopapain on , intervertebral discs of dogs. Clinical Orthopaedics 921: 281-295 Hayes A J, Benjamin M, Ralphs J R 2001 Extracellular matrix in development of the intervertebral disc. Matrix Biology 20: 107-121 Higuchi M, Abe K, Kaneda K 1983 Changes in the nucleus pulposus of the intervertebral disc in bipedal mice: a light and electron microscopic study. Clinical Orthopaedics 175: 251-257 Holm S, Nachemson A 1982 Nutritional changes in the canine intervertebral disc after spinal fusion. Clinical Orthopaedics 169: 243-258 Holm S, Nachemson A 1983 Variation in the nutrition of the canine intervertebral disc induced by motion. Spine 8: 866-874. Holm S, Nachemson A 1988 Nutrition of the intervertebral disc: acute effects of cigarette smoking: an experimental animal study. Upsala Journal of Medical Sciences 93: 91-99 Holm S, Maroudas A, Urban J P, Selstam G, Nachemson A 1981 Nutrition of the intervertebral disc: solute transport and metabolism. Connective Tissue Research 8: 101-119 Horner H A, Urban J P 2001 2001 Volvo Award winner in basic science studies. Effect of nutrient supply on the viability of cells from the nucleus pulposus of the intervertebral disc. Spine 26: 2543-2549 Hughes C E, Little C B, Buttner F H, Bartnik E, Caterson B 1998 Differential expression of aggrecanase and matrix metalloproteinase activity in chondrocytes isolated from bovine and porcine articular cartilage. Journal of Biological Chemistry 273: 30576-30582 latridis J C, Kumar S, Foster R J, Weidenbaum M, Mow V C 1999 Shear mechanical properties of human lumbar annulus fibrosus. Journal of Orthopaedic Research 17: 732-737 Indahl A, Kaigle A M, Reikeras 0, Holm S H 1997 Interaction between the porcine lumbar intervertebral disc, zygapophysial joints, and paraspinal muscles. Spine 22: 2834-2840 Iozzo R V Murdoch A 0 1996 Proteoglycans of the extracellular , environment: clues from the gene and protein side offer novel perspectives in molecular diversity and function. FASEB Journal 10: 59�14 Ishihara H, Urban J P 1999 Effects of low oxygen concentrations and metabolic inhibitors on proteoglycan and protein synthesis rates in the intervertebral disc. Journal of Orthopaedic Research 17: 829-835 Ishihara H, McNally 0 S, Urban J G, Hall A C 1996 Effects of hydrostatic pressure on matrix synthesis in different regions of the intervertebral disk. Journal of Applied PhYSiology 80: 839-846 Johnstone B, Bayliss M T 1995 The large proteoglycans of the human intervertebral disc: changes in their biosynthesis and structure with age, topography, and pathology. Spine 20: 674-684 Johnstone B, Markopolous M, Neame P, Caterson B 1993 Identification and characterization of glycanated and non-glycanated forms of biglycan and decorin in the human intervertebral disc. Biochemical Journal 292: 661-666 Jones J P 1997 Subchondral osteonecrosis can conceivably cause disk degeneration and 'primary'osteoarthritis. In: Urbaniak J R, Jones J P (eds) Osteonecrosis. American Academy of Orthopaedic Surgery, Park Ridge, Illinois, pp 135-142

Chemistry of t h e intervertebral disc in relation to functional requ irements


human lumbar intervertebral discs. Acta Orthopaedica Scandinavica 41 : 589--607 Nerlich A G, Schleicher E D, Boos N 1997 1997 Volvo Award winner in ' basic science studies. Immunohistologic markers for age-related changes of human lumbar intervertebral discs. Spine 22: 2781-2795 Oegema T R J, Bradford D S, Cooper K M, Hunter R E 1983 Comparison of the biochemistry of proteoglycans isolated from normal, idiopathic scoliotic and cerebral palsy spines. Spine 8: 378-384 Oegema T R, Johnson S L, Aguiar D J, Ogilvie J W 2000 Fibronectin and its fragments increase with degeneration in the human intervertebral disc. Spine 25: 2742-2747 O'Hara B P, Urban J P G, Maroudas A 1990 Influence of cyclic loading on the nutrition of articular cartilage. Annals of the Rheumatic Diseases 49: 536-539 Ohshima H, Urban J P G, Bergel D H 1995 The effect of static load on matrix synthesis rates in the intervertebral disc measured in vitro by a new perfusion technique. Journal of Orthopaedic Research 13: 22-29 Osada R, Ohshima H, Ishihara H et al 1996 Autocrine/paracrine mechanism of insulin-like growth factor-1 secretion, and the effect of insulin-like growth factor-I on proteoglycan synthesis in bovine intervertebral discs. Journal of Orthopaedic Research 14: 690--699 Paajanen H, Lehto I, Alanen A, Erkintalo M, Komu M 1994 Diurnal fluid changes of lumbar discs measured indirectly by magnetic resonance imaging. Journal of Orthopaedic Research 12: 509-514 Paassilta P, Lohiniva J, Goring H H et al 2001 Identification of a novel common genetic risk factor for lumbar disk disease. JAMA 285: 1843-1849 Pazzaglia U E, Andrini L, Di Nucci A 1997 The effects of mechanical forces on bones and joints: experimental study on the rat taiL Journal of Bone and Joint Surgery (British volume) 79: 1024-1030 Peacock A 1951 Observations on the postnatal development of the intervertebral disc in man. Journal of Anatomy 86: 1 62-179 Puus�arvi K, Lammi M, Kiviranta I, Helminen H J, Tammi M 1993 Proteoglycan syntheSiS in canine intervertebral discs after long distance running training. Journal of Orthopaedic Research 11: 738-746 Roberts S, Menage J, Urban J P G 1989 Biochemical and structural properties of the cartilage end-plate and its relation to the intervertebral disc. Spine 14: 166-174 Roberts S, Menage J, Eisenstein S M 1993 The cartilage end-plate and intervertebral disc in scoliosis: calcification and other sequelae. Journal of Orthopaedic Research 11: 747-757 Roberts S, Eisenstein S M, Menage J, Evans E H, Ashton I K 1995 Mechanoreceptors in intervertebral discs: morphology, distribution, and neuropeptides [see comments] . Spine 20: 2645-2651 Roberts S, Urban J P G, Evans H, Eisenstein S M 1996 Transport properties of the human cartilage endplate in relation to its composition and calcification. Spine 21: 415-420 Roberts S, Caterson B, Menage J, Evans E H, Jaffray D C, Eisenstein S M 2000 Matrix metalloproteinases and aggrecanase: their role in disorders of the human intervertebral disc. Spine 25: 3005-3013 Sambrook P N, MacGregor A J, Spector T D 1999 Genetic influences on cervical and lumbar disc degeneration: a magnetic resonance imaging study in twins. Arthritis and Rheumatism 42: 366-372

Sato K, Kikuchi S, Yonezawa T 1999 In vivo intradiscal pressure measurement in healthy individuals and in patients with ongoing back problems. Spine 24: 2468-2474 Schneiderman R, Rosenberg N, Hiss J et al 1995 Concentration and size distribution of insulin-like growth factor-I in human normal and osteoarthritic synovial fluid and cartilage. Archives of Biochemistry and Biophysics 324: 173-188 Sedowofia K A, Tomlinson I W, Weiss J B 1982 Collagenlytic enzyme systems in human intervertebral discs: their control mechanism, and their possible roles in the initiation of biomechanical failure. Spine 7: 213-222 Shirazi-Adl S A, Shrivastava S C, Ahmed A M 1984 Stress analysis of the lumbar disc-body unit in compression: a three-dimensional non­ linear finite element study. Spine 9: 120-134 Sztrolovics R, Alini M, Mort J S, Roughley P J 1999 Age-related changes in fibromodulin and lurnican in human intervertebral discs. Spine 24: 1765-1771 Takeda T 1975 Three-dimensional observations of collagen framework of human lumbar discs. Journal of the Japanese Orthopedic Association 49: 45-57 Thomas R D M, Batten J J, Want S, McCarthy I D, Brown M, Hughes S P F 1995 A new in-vitro model to investigate antibiotic penetration of the intervertebral disc. Journal of Bone and Joint Surgery 77B: 967-970 Thompson J P, Pearce R H, Schechter M T, Adams M E, Tsang I K Y, Bishop P B 1990 Preliminary evaluation of a scheme for grading the gross morphology of the human intervertebral disc. Spine 15: 411-415 Thompson J P, Oegema T R, Bradford D S 1991 Stimulation of mature canine intervertebral disc by growth factors. Spine 16: 253-260 Thornton W E, Moore T P, Pool S L 1987 Fluid shifts in weightlessness. Aviation, Space and Environmental Medicine 58: A86-A90 Urban J P, Maroudas A 1981 Swelling of tlle intervertebral disc in vitro. Connective Tissue Research 9: 1-10 Urban J P G, Roberts S 1995 Development and degeneration of the intervertebral discs. Molecular Medicine Today 1 : 329-335 Urban J P G, Holm S, Maroudas A 1979 Diffusion of small solutes into the intervertebral disc: an in vivo study. Biorheology 15: 203-223 Urban M R, Fairbank J C, Etherington P J, Loh F L, Winlove C P, Urban J P 2001a Electrochemical measurement of transport into scoliotic intervertebral discs in vivo using nitrous oxide as a tracer. Spine 26: 984-990 Urban M R, Fairbank J C, Bibby S R, Urban J P 2001b Intervertebral disc composition in neuromuscular scoliosis: changes in cell density and glycosaminoglycan concentration at the curve apex. Spine 26: 610--617 Videman T, Leppavuori J, Kaprio J 1998 latragenic polymorphisms of the vitamin D receptor gene associated with intervertebral disc degeneration. Proceedings of the International Society for the Study of the Lumbar Spine, Brussels, 59 [abstract] W ilke H J, Neef P, Cainli M, Hoogland T, Claes L E 1999 New in vivo measurements of pressures in the intervertebral disc in daily life. Spine 24: 755-762 Woods A, Couchman J R 2001 Syndecan-4 and focal adhesion function. Current Opinion in Cell Biology 13: 578-583





CI'inical biomechanics of the thoracic spine including the ribcage
s. J. Ed mondston


55 56 56 57 58

Loadbearing biomechanics of the thoracic spine Biomechanics of the thoracic kyphosis Regional mobility of the thoracic spine
Upper thoracic region Mid-thoracic region Low thoracic region flexion 60 61 58 58 59

Mechanical stability of the thoracic spine

Movement coupling in rotation/lateral Movement of the thoracic spine and ribcage during respiration

Muscle actions on the thoracic spine and ribcage practice Conclusion


Biomechanical considerations in manual therapy


An understanding of the biomechanics of the thoracic spine and ribcage is important in the practice of manual therapy as it provides a basis for the interpretation of patterns of clinical presentation in patients with 'mechanical' pain dis­ orders of the thoracic region, Surprisingly, the thoracic spine has been a relatively limited focus of biomechanical research which may explain why this region of the spine has been considered an enigma relative to the cervical and lumbar regions (Singer & Edmondston 2000), The percep­ tion that thoracic musculoskeletal pain disorders are less common is supported by the limited epidemiological data which suggests that these account for less than 15% of spinal pain presentations in the general population and in manual therapy practice (Hinkley & Drysdale 1995, Linton et aI1998). Despite this, the severity of symptoms and asso­ ciated level of disability can be equal to those of patients with lumbar spine disorders (Occhipiniti et a11993), which may explain the resurgent interest in this region of the spine from a clinical and biomechanical perspective. The presence of the ribcage and the complex mechanical interaction between the spine and ribcage present signifi­ cant methodological problems for biomechanical studies of the thoracic spine. Finite element models and animal labo­ ratory studies provide much of the data on movement pat­ terns and stability of the thoracic spine/ribcage complex. Many of the ex vivo studies of human thoracic mechanics have been conducted on specimens without an intact ribcage which may limit the applicability of the findings to clinical practice as thoracic spine mobility and loadbearing are significantly influenced by the ribcage (Andriacchi et al 1974, Berg 1993). Clinical studies of thoracic posture mechanics have used radiological imaging techniques (Goh et al 2000) but these techniques have limited value in kine­ matic studies. However, a clearer understanding of thoracic spine mechanics has been achieved though the combined results of motion analysis studies of asymptomatic subjects in conjunction with clinical observation (Gregerson & Lucas 1967, Lee 1994, Willems et aI1996).



Although the thoracic spine is anatomically well defined, the functional boundaries of this region of the spine are less distinct. In this review, emphasis is given to regional differ­ ences in the mechanics of the thoracic spine, which are reflected in the skeletal and articular anatomy. The upper thoracic spine may be considered as being functionally part of the cervical spine, while the low thoracic motion seg­ ment anatomy results in movement patterns more closely resembling those of the lumbar spine. The 'functional' tho­ racic spine therefore seems to consist of the motion seg­ ments between T3 and T9 (Lee 1993). This arbitrary division of the thoracic spine into functional regions seems consis­ tent with the patterns of clinical presentation of mechanical pain disorders which have been described in this region (Lee 1994, Singer & Edmondston 2000). The biomechanics of the thoracic spine will be consid­ ered in this review in relation to the two common patterns of clinical presentation. The first is the disorders where pain is predominantly associated with spinal loading and load attenuation. The biomechanics of thoracic loadbearing are reviewed with reference to loadsharing in the motion seg­ ment, influences on spinal curvature and the muscular and postural responses to loadbearing. The second issue relates to situations where symptoms relate more to movement activities or restriction of movement. The interaction between mechanical stability and mobility requirements is reviewed, with reference to the variability in the range and patterns of movement in the different regions of the tho­ racic spine. The primary objective is to summarize the cur­ rent knowledge of thoracic spine and ribcage biomechanics which have particular relevance to the practice of manual therapy.

The compressive load on the thoracic spine increases cau­ dally from about 9% of body weight at T1 to 47% of body weight at T12 (White 1969). The ability to sustain the increasing loading demands is achieved through a progres­ sive increase in vertebral body size, end-plate cross­ sectional area and bone content, particularly in the lower six vertebral segments (Edmondston et a11994a, Singer et al 1995). Cancellous bone density and architecture is relatively constant between T2 and T12 which suggests that the skele­ tal adaptation to increasing load is that of an increase in bone mass rather than of cancellous bone density (Edmondston et aI1994b). The loadbearing capacity of the thoracic spine may be up to three times greater when the ribcage is intact (Andriacchi et al 1974). According to Pal & Routal (1986), 76% of compressive load in the upper tho­ racic spine is transferred through the vertebral body I inter­ vertebral disc complex. This loadsharing ratio is likely to be similar in the mid-thoracic region, due to the anterior loca­ tion of the line of gravity relative to the spine. The prefer­ ential loading of the anterior spinal structures in the

mid-thoracic region is reflected in the higher incidence of disc degeneration and vertebral body deformity in these segments (Singer 1997, Goh et aI1999). In the low thoracic spine, a greater proportion of the compressive load may be transferred through the posterior column formed by the interlocking laminae and articular facets, as well as the lower costovertebral joints (Pal & RoutaI1987). The medial taper of the articular facets and 'wrap-around' configura­ tion of the mortice joints at the thoracolumbar junction would act to provide a stable platform for compressive loadbearing in this region of the spine, while restricting tor­ sional mobility (Singer & Malmivaara 2000). The intervertebral disc has an important role in attenuat­ ing the static and impact compressive loads applied to the thoracic spine during functional and recreational activities. Although the response of the lumbar disc to compressive loading has been investigated in radiological and labora­ tory studies, there are few comparative studies of the mechanical properties of thoracic discs in compression (Martinez et al 1997, Wisleder et al 2001). Regional varia­ tions in the mechanical properties of thoracic intervertebral discs in response to compressive loading have been reported in ex vivo studies. When normalized for differ­ ences in height, the upper and mid-thoracic discs undergo greater deformation and creep in response to a specific load than do the discs in the low thoracic and upper lumbar regions (Koeller et aI1984). Differences in water content do not appear to account for the more viscous mechanical behaviour of the upper and mid-thoracic discs in response to compressive load (Koeller et aI1984). Instead this may be due to differences in disc morphology and biochemistry, and the structural arrangement of the annular lamellae (Pooni et al 1986, Scott et al 1994, Putz & Miiller-Gerbl 2000). In the lumbar spine, compressive load is evenly dis­ tributed across the surface of the vertebral end-plate, inde­ pendent of the position of the motion segment. In the thoracic spine, the uniform load distribution across the end­ plate becomes asymmetric when loaded outside the neutral position (Horst & Brinckmann 1981). Since the thoracic discs are a potential source of pain, these observations in relation to the biomechanical response to compressive loading may explain the common clinical presentation of mid-thoracic pain associated with sustained loading activities such as word processing and driving. Indeed, a higher prevalence of thoracic pain has been reported in an occupational survey comparing spinal pain symptoms in bus drivers (28%) compared to employ­ ees in the same company with non-driving occupations (10%) (Anderson 1992).

The thoracic kyphosis is the primary curve of the spinal axis, persisting from embryological development. In stand­ ing postures the form of the thoracic spine is maintained by the tensile forces in the posterior ligaments and spinal

Clinical biomechanics of the thoracic spine including the ribcage


extensor muscles, and the balanced compressive loads transferred through the vertebral bodies and discs (White et aI1977). The thoracic curvature in standing is largely influ­ enced by the location of the line of gravity and the shape of the vertebral bodies and intervertebral discs (Pearsall & Reid 1992, Manns et al 1996, Goh et aI1999). In a compari­ son of clinical and post mortem radiographs, Singer et al (1994) found little difference in the resting form of the kyphosis confirming the importance of ligamentous ten­ sion and skeletal and disc morphology in determining tho­ racic curvature. The resting length of antagonistic muscle groups and the level of recruitment of trunk musculature have been hypothesized to influence the sagittal plane curvatures of the spine (White & Sahrmann 1994). However, Toppenberg & Bullock (1986) were unable to demonstrate an association between the length of trunk and lower limb muscles and the thoracic kyphosis. In relaxed standing, relatively low levels of phasic muscle activity are required to maintain the upright posture and correct for postural sway (Ortengren & Andersson 1977). This low-level muscle activity would seem unlikely to have much influence on thoracic curva­ ture. Similarly, trunk muscle strength is unlikely to influ­ ence neutral spinal curvature, a hypothesis confirmed by Walker et al (1987). Incremental spinal loading studies have examined the influence of trunk muscle recruitment on tho­ racic curvature. Klausen (1965) observed no change in tho­ racic curvature when external loads of up to 40 kg were applied using a backpack. Similarly, Edmondston et al (2000) reported no change in the thoracic kyphosis, despite a linear increase in EMG activity of the erector spinae mus­ cles, when the subjects held loads of up to 20% of body weight. A non-linear increase in abdominal muscle recruit­ ment was also noted during this loading study. Hence the optimal response to loading in the thoracic spine appears to be one in which the neutral curvature is maintained through an increase in the balanced trunk muscle activation associated with unloaded standing.

Normal mechanical function of the thoracic spine is dependent on an appropriate interaction between mobility and stability in the motion segments. The ribcage and ster­ num provide additional stability for the thoracic spine dur­ ing loadbearing and movement, and thoracic stiffness is significantly reduced when the integrity of the ribcage is compromised (Berg 1993, Shea et al 1996). Stability during dynamic loading tasks is further enhanced by an increase in intrathoracic pressure, which is achieved through coordi­ nated contraction of the diaphragm, together with the deep abdominal and intercostal muscles (Morris et al 1961, Hodges & Gandevia 2000). In response to an applied force, the motion segment dis­ plays non-linear behaviour, with minimal resistance to movement initially (neutral zone), followed by an elastic

zone in which movement (displacement) is proportional to load (Panjabi et aI1989). Control of segmental movement in the neutral zone is dependent on muscle contraction while in the elastic range motion control is provided by ligamen­ tous tension and the intervertebral disc (Panjabi 1992). In the lumbar spine, the range of the neutral zone is greatest in the sagittal plane while in the thoracic spine the sagittal plane neutral zone is smaller than in the coronal and hori­ zontal planes (Oda et a11996) (Table 6.1). It is evident from experimental studies that considerable anatomical disruption is required to produce mechanical instability in the thoracic spine. Transection of all posterior ligaments and destabilization of the costovertebral joint is required to cause flexion instability of the motion segment (Shea et aI1996). Similarly, extension stability in the motion segment is compromised following complete transection of the intervertebral disc and rib head resection (Panjabi et al 1981, Feiertag et al 1995). Stability of the thoracic spine in the coronal plane is dependent more on the costotransverse ligament complex than the midline ligaments. The strain in the lateral ligaments of the thoracic spine may be up to 5.6% with only 1 degree of lateral flexion while the strain in the midline ligaments, for the equivalent movement, has been shown to be only 1 % (Panjabi & Goel 1982, Jiang et aI1994). The influence of the posterior ligaments and rib joints on the mobility and neutral zone of the thoracic motion seg­ ments was examined by Oda and co-workers (1996) using a canine model. Following removal of these structures, the neutral zone increased by less than 2 degrees and 4 degrees in the sagittal and axial planes respectively. The greatest increase in neutral zone was in the frontal plane where the change was 7.3 degrees. The changes in the neutral zone of the motion segment may result from injury or degeneration of the motion seg­ ment, particularly of the intervertebral disc. In the lumbar spine, changes in neutral zone, which may relate to clinical instability, are greater in the sagittal plane (Wilke et aI1995). Similarly, radiological and clinical patterns of lumbar seg­ mental instability are observed more commonly with sagit­ tal plane movements (Boden & Wiesel 1990, O'Sullivan 2000). In contrast, the sagittal plane neutral zone in the tho­ racic spine is very small due to the narrow disc height and coronal orientation of the zygapophysial joints, which

Table 6.1

Comparison of neutral zone ranges for thoracic

and lumbar spine motion segments
Sagittal plane Thoracic* Lumbar' Coronal plane Axial plane

0.6 1.7

3.5 2.9

2. 1 0.2

All numbers are in degrees. Data from ·Oda et al 1996, +Wilke et al 1995.



strongly constrain sagittal movement. The higher range of unconstrained movement (neutral zone) in axial rotation and lateral flexion is consistent with the description of rota­ tional instability as a pattern of patient presentation in the mid-thoracic spine (Lee 1994). While motion palpation tests for examining the stability of the thoracic motion segments have been proposed (Lowcock 1991), it is not possible to examine the range or patterns of segmental motion in the thoracic spine using radiological imaging techniques.

Normal movement of the thoracic spine is required to facil­ itate functional tasks and recreational activities. An under­ standing of the kinematics of the thoracic spine, including regional variations and the anatomical influences on move­ ment, is required in the interpretation of any movement examination of patients with thoracic pain disorders. Unfortunately, the unique anatomy of the thoracic spine, particularly the presence of the ribcage, presents significant difficulties for in vivo investigations of thoracic movement. Much of the reported data come from cadaveric studies which are limited by the requirement to dissect the ribcage and related muscles prior to analysis. Stereo-radiography techniques cannot be used in the thoracic spine due to poor vertebral definition and superimposition of the ribs, although rotational mobility has been measured using CT (Singer et al 1989). Given the ethical constraints associated with invasive measurement techniques, surface measure­ ments using electromagnetic motion analysis systems are increasingly being employed (Willems et al 1996). However, the extent to which surface measurements reflect the movement patterns of the underlying joints remains questionable (Stokes 2000). Despite these difficulties, data derived from studies using each of these analysis tech­ niques provide a more complete understanding of the kine­ matics of the thoracic spine and support the development of biomechanical models.

rib moves inferiorly, occurs during extension of the upper thoracic spine (Lee 1993). The kinematics of upper thoracic rotation and lateral flexion are more complex due to the asymmetrical move­ ment patterns in the spinal motion segments and ribs. The constraining influence of the ribs on these movements is confirmed by the overall lower ranges of segmental motion reported in ex vivo studies compared to measurements from human subjects. Lateral flexion occurs around an axis located within the disc space between the mid-disc and ipsilateral margin of the vertebra (White 1969). The bilateral range of upper thoracic lateral flexion reported in ex vivo studies is 6- 8 degrees per segment compared to about 4 degrees per segment from clinical studies (White 1969, Willems et al 1996). Axial rotation in the upper thoracic spine occurs around an axis located forward of the anterior margin of the vertebral body (Davis 1959). The in vivo range of upper thoracic rotation is about 8 degrees per seg­ ment compared to about 12 degrees per segment from the ex vivo studies (White 1969, Willems et aI1996). Movement coupling between rotation and lateral flexion in this region of the spine may be inconsistent within and between indi­ viduals due to the influence of the muscles which span the cervicothoracic junction, and the associated effect on spinal and rib movement (Willems et al 1996). Descriptions of rib movement associated with coronal and axial plane spinal movement are based on clinical observation (Lee 1994). Lateral flexion of the upper thoracic spine is associated with ipsilateral anterior rotation and contralateral posterior rotation of the upper ribs. Rib movement is more pro­ nounced during cervicothoracic rotation where posterior rotation of the right ribs and anterior rotation of the left ribs accompanies right rotation and vice versa.

Mid-thoracic region
The mid-thoracic spine (T3-9) is most mobile in axial rota­ tion with the range of movement achievable in sitting being the same as that in standing (Gregersen & Lucas 1967). The axis of rotation for this movement lies within the vertebral body, which, together with the coronal plane orientation of the zygapophysial joints, promotes lateral translation of the articular facets. This is accompanied by ipsilateral transla­ tion and tilt of the vertebral body. However, these coupled movements would be limited due to the thin intervertebral discs and tension in the costal ligaments (Davis 1959). Tension developed in the costal ligaments causes posterior rotation of the ipsilateral rib and anterior rotation of the contralateral rib during axial rotation in the mid-thoracic spine (Lee 1993) (Fig. 6.1). The range of axial rotation in the mid-thoracic spine has been reported as being about 10 degrees per segment, based on cadaveric and in vivo sur­ . face measurements (White 1969, Willems et aI1996). in con­ trast, Gregersen & Lucas (1967) were able to obtain more direct measurement from human subjects by recording movements from Steinmann pins inserted into the spinous

Upper thoracic region
Descriptions of the ranges of movement in the thoracic spine highlight the regional differences in motion segment anatomy. Upper thoracic mobility contributes to normal cervical spine function and to functional movements of the thorax. Sagittal movements are accompanied by little movement in the other planes, possibly due to the symmet­ rical anterior rotation of the upper ribs which may act to constrain coupled movements (Willems et aI1996). A range of upper thoracic sagittal movement of about 5 degrees per segment has been reported in both in vivo and cadaveric studies (White 1969, Willems et aI1996). The proportion of this range which is extension is reported as being between 30 and 50% which may reflect differences in the reference point for measurement in these studies. Symmetrical poste­ rior rotation of the ribs, such that the posterior part of the

Clinical biomechanics of the thoracic spine including the ribcage


Sagittal plane movement is relatively limited in the mid­ thoracic spine. The axis of rotation for sagittal rotation is located in the disc space of the caudad motion segment. However, the exact location is different for flexion and extension (Panjabi et al 1984). Anterior sagittal rotation (flexion) and the associated anterior translation are con­ strained by the vertical articular facets of the zygapophysial joints (Panjabi et al 1984). Flexion is limited by tension in the posterior spinal ligaments and approximation of the ribs, which rotate anteriorly during this movement (Lee 1993). Mid-thoracic extension is associated with posterior translation of the superior vertebra, which is less con­ strained by the articular facets of the zygapophysial jOints. In contrast, vertebral motion in extension is guided by the contact of the inferior articular facet or the spinous process on the vertebra below resulting in a constrained axis of rotation (Panjabi et aI1984). The posterior vertebral transla­ tion during extension induces posterior rotation of the ribs (Lee 1993). The range of sagittal movement has been deter­ mined as being about 5 degrees per segment in cadaveric and in vivo studies (White 1969, Willems et al 1996). The consistency between cadaveric and clinical studies is possi­ bly due to the greater influence of the zygapophysial joints, rather than the ribcage, in determining the range of sagittal movement.

Low thoracic region
Response of the mid-thoracic spine to rotation. Right rotation of the trunk is associated with ipsilateral lateral flexion of the thoracic spine. Right rotation of the thorax is associated with posterior rotation of the ipsilateral ribs and anterior rotation of the contralateral ribs (white arrows) .
Figure 6.1

processes. These investigators reported a segmental range of axial rotation of about 5 degrees per segment. During normal gait, axial rotation is greatest in the mid-thoracic segments (up to 2.5 degrees per segment) (Gregersen & Lucas 1967). The greater rotational mobility of the mid-tho­ racic spine, and the associated torsion and shear forces transferred to the intervertebral discs, may contribute to the higher prevalence of disc degeneration in these segments (Singer 1997). Approximation of the ribcage during lateral flexion of the mid-thoracic spine limits mobility in the coronal plane. A segmental range of lateral flexion of 4 degrees per seg­ ment has been reported in clinical studies compared to 6 degrees per segment in the cadaveric experiments where the ribcage was removed (White 1969, Willems et aI1996). Lateral flexion of the mid-thoracic spine produces concur­ rent anterior rotation of the ipsilateral ribs and posterior rotation of the contralateral ribs. This asymmetrical rib movement may contribute the contralateral rotation of the thorax which is observed clinically during trunk lateral flexion (Lee 1993).

Movement in the low thoracic spine is influenced by the variability in posterior element morphology and the anatomy of the lower two ribs, which articulate with one vertebral body and have no anterior attachment. Zygapophysial joint asymmetry (tropism) and different patterns of transition from coronal to sagittal orientation result in considerable variability between individuals in the ranges of motion and patterns of coupled motion in this region (Gregersen & Lucas 1967, Singer et al 1989). The greater disc height and more sagittal orientation of the zygapophysial joints in the low thoracic region facilitate mobility in the sagittal plane (White 1969, Pooni et al 1986). Evidence for these anatomical influences on movement in the low thoracic spine comes from the cadaveric study of White (1969) who reported 8 degrees of sagittal movement at T9/1O compared with 20 degrees at Tll/12. This com­ pares with 5 degrees per segment between T8 and T12 determined using a surface measurement technique in an in vivo study (Willems et al 1996). Mobility in the coronal plane in the low thoracic region is similar to that in the upper and mid-thoracic segments. A range of 6 degrees per segment between T8 and T12 was reported in the clinical study of Willems and co-workers (1996). In contrast, cadav­ eric measurements of low thoracic lateral flexion show an increase in range from 6 degrees at T9/1O to 12 degrees at Tll/12 (White 1969). These results highlight the influ­ ence of zygapophysial joint orientation on mobility in the thoracolumbar junction region (T11-Ll) compared to the



adjacent cephalad segments (Malmivaara et al 1987, Singer et aI1989). The low thoracic spine (T8-12) has a more limited range of axial rotation compared to the upper and mid-thoracic regions. In vivo studies have reported ranges of motion of between 5 and 7 degrees per segment (Gregersen & Lucas 1967, Willems et al 1996). As with movement in the other planes, variability in the segmental range of axial rotation within this region is due to the changing orientation of the zygapophysial joints. Based on measurements from CT scans, unilateral segmental rotation was found to decrease from 2.8 degrees per segment at TlO/ll to 1.8 degrees per segment at Tl2/Ll (Singer et al 1989). These investigators found no significant difference in segmental rotation between subjects with an abrupt change in zygapophysial joint orientation compared to those in which it was more gradual. However, a 'mortice'-type configuration of the zygapophysial joints observed in some individuals in this region may further constrain axial rotation due to the medial taper of the joint surfaces and the extended mamil­ lary process of the superior articular facet (Singer et al 1989).

Movement coupling in rotation/lateral flexion
Movement of the thoracic spine rarely occurs in a single plane. Due to various structural and anatomical influences, spinal movement in one plane is inevitably accompanied by one or more coupled movements (Harrison et al 1998). Movement coupling principles provide the foundation for the interpretation of patterns of movement impairment and technique selection in some methods of manual therapy practice (Evjenth & Hamberg 1984, Gibbons & Tehan 1998). In particular, patterns of movement coupling in the frontal

and horizontal planes have been the focus of numerous cadaveric and in vivo studies for almost 100 years (Lovett 1905). It is apparent that the primary direction of movement influences the range and direction of coupled movements and that regional differences in coupled motion exist in the thoracic spine (Willems et al 1996). These regional varia­ tions in movement coupling may be due to vertebral orien­ tation within the kyphosis, zygapophysial joint anatomy and the costal articulations (Veldhuizen & Scholten 1987, Singer et aI1989). A summary of studies examining coupled rotation and lateral flexion is presented in Table 6.2. The variation in movement coupling between rotation and lateral flexion is likely to be due to differences in study design and meas­ urement techniques (Gregersen & Lucas 1967, Panjabi et al 1976, Willems et al 1996). Furthermore, analysis of coupled movement is difficult owing to the small ranges of move­ ment which are subject to significant measurement error (Panjabi et al 1976). In these studies, measurements have been derived from the spine rather than from analysis of movement of the thorax (spine and ribcage) as a whole complex. This seems important considering the interaction between spinal and rib movement as previously described. From clinical observation it does appear that rotation is associated with coupled ipsilateral lateral flexion of the spine (see Fig. 6.1). However, consideration of the associ­ ated rib movement leads to the (untested) hypothesis that rotation of the thorax (spine and ribcage) is associated with contralateral lateral flexion and vice versa. Movement of the thorax into right rotation is associated with posterior rotation of the ipsilateral ribs (Lee 1993). In contrast, right lateral flexion is associated with anterior rotation of the ipsilateral ribs. Therefore, it seems likely that right rotation of the thorax would be accompanied by coupled left lateral

Table 6.2

Summary of studies which have examined patterns of coupled movement in the thoracic spine

Author Gregersen Et Lucas 1967 White 1969 Panjabi et al 1976 Lee 1993 Willems et al 1996

Method Normal volunteers Cadaver Cadaver Biomechanical model Surface measurement (3-Space Fastrak system)

Region Upper and middle Lower Upper Middle and lower Middle Middle Middle Upper Middle Lower

Primary movement LF LF LF LF Rot. LF Rot. LF LF Rot. LF Rot. LF Rot.

Coupled movement Ipsilateral rot. Variable pattern Ipsilateral rot. Ipsilateral rot: (variable) Contralateral LF Contralateral rot. Ipsilateral LF Contralateral rot. Contralateral rot. (53%) Contralateral LF (82%) Ipsilateral rot. (83%) Ipsilateral LF (99%) Ipsilateral rot. (680Jo) Ipsilateral LF (93%)



lateral flexion; rot.



Clinical biomechanics of the thoracic spine including the ribcage


flexion as in both cases the right-sided ribs would rotate posteriorly. This functional approach to the interpretation of movement coupling in the thoracic spine/ribcage com­ plex appears to have greater relevance to the practice of manual therapy than consideration of coupled motion of the spine in isolation.

Movement of the thoracic spine and ribcage during respiration
Movement of the ribcage during inspiration is initiated by the diaphragm, which elevates the lower ribs as the con­ traction causes depression of the central tendon (DeTroyer & Estenne 1988). Rib movement occurs around a mediolat­ eraI axis, which extends from the costovertebral joint towards the rib tubercle (Rickenbacher et al 1985, Saumarez 1986). In the upper ribs this axis is located at about 35 degrees to the coronal plane whereas in the lower ribs the axis is oriented closer to the sagittal plane. Consequently, movement of the upper ribs elevates the sternum and increases the anteroposterior diameter of the ribcage ('pump-handle') while movement of the lower ribs has a greater influence on the lateral dimensions of the ribcage (,bucket-handle') (Harris & Holmes 1996). Although both actions of the ribs occur simultaneously, the proportion of 'pump-handle' movement is greater in the upper ribs while the 'bucket-handle' action is more dominant in the lower ribs (Mitchell & Mitchell 1995). The lower two ribs have no anterior attachment and have a 'caliper ' -type action (Mitchell & Mitchell 1995). During quiet respiration there is relatively little move­ ment of the upper ribs. However, on exertion, upper ribcage movement increases due to the action of the acces­ sory respiratory muscles (scalenii, sternomastoid and pec­ toralis minor) (DeTroyer & Estenne 1988). The role of the intercostal muscles in respiration remains contentious but these muscles could have an inspiratory or expiratory func­ tion dependent on their level of activity in different costal segments (Loring & Woodbridge 1991). Deep inspiration in sitting is associated with extension of the lumbar and tho­ racic spine, possibly to accommodate the concurrent poste­ rior (pump-handle) rotation of the ribs (Leong et aI1999).

extensor moment, reducing the tension generated in the extensor muscles and the associated compressive forces transferred to the thoracolumbar spine (Morris et al 1961, Daggfeldt & Thorstensson 1997). Generation of axial torque provides trunk rotation dur­ ing locomotion, and for sporting activities such as golf and racquet sports. The oblique abdominal muscles generate the forces required for thoracic spine rotation. Due to the anterior location of these muscles, contraction is associated with combined flexion and rotation of the trunk (Bogduk 1986). The flexion movement is resisted by simultaneous contraction of the ipsilateral thoracic fibres of iliocostalis and longissumus (Rickenbacher et al 1985). More specific control of thoracic rotation may be achieved through uni­ lateral contraction of the contralateral thoracic multifidus and rotatores muscles. The oblique orientation of these fibres promotes movement in the horizontal plane rather than the extension movement generated by the lumbar multifidus (Bojadsen et al 2000). The relative role of the oblique abdominal and thoracic erector spinae in generat­ ing axial torque in the thoracic spine remains uncertain. Lateral flexion of the thorax is controlled by the eccentric action of iliocostalis and longissumus, with a lesser contri­ bution from the medial intersegmental muscles. The con­ tralateral medial tract muscles (semispinalis, multifidus and rotatores) control the associated rotation produced by the long fibres of iliocostalis. (Rickenbacher et aI1985).

Movement of the thoracic spine and ribcage is dependent on coordinated contraction of the associated musculature. Sagittal movements of the thorax are achieved through the activation of the thoracic fibres of iliocostalis and longis­ simus, which act around the thoracic kyphosis (Macintosh & Bogduk 1994). Generation of extension moments during functional tasks is associated with synergistic activation of the diaphragm and abdominal muscles, which elevate intra-abdominal pressure (lAP) (Morris et al 1961, Stokes 2000). The increase in lAP in particular contributes to the

Knowledge of the regional biomechanics of the thoracic spine and ribcage assists the clinician in the interpretation of active movement and motion palpation examination in relation to the patient's symptoms. Normal mechanics of the cervical spine and shoulder are dependent on normal mobility in the upper thoracic spine. A habitual flexed upper thoracic posture may reduce the capacity of the mus­ cles, which provide cervicothoracic retraction to work in the functional range. Further, the upper ribs will be drawn into anterior rotation due to the flexed position of the upper thoracic spine. Restriction of cervical extension and rotation movements is inevitable due to the restriction of upper rib mobility and the requirement for movement out of the neu­ tral spinal alignment. Consequently, restricted upper tho­ racic mobility may increase the movement demands on the more mobile lower cervical segments, with the potential for symptom development or exacerbation. Upper thoracic extension is required to accommodate the later range of bilateral flexion of the shoulders, while ipsi­ lateral flexion of the upper thoracic spine is observed dur­ ing unilateral shoulder elevation (Culham & Peat 1993, Sobel et aI1996). Consequently, changes in upper thoracic posture and mobility may lead to subacromial pathology due to the effects on scapula and glenohumeral mechanics (Sobel et aI1996). Similarly, restriction of upper rib mobility



may produce symptoms and physical signs consistent with those of subacromial impingement or thoracic outlet syn­ drome (Lindgren & Leino1988, Boyle 1999). Based on these observations, examination of upper thoracic and rib mobil­ ity would be important in patients with shoulder pain related to overhead activities. Due to their location in the apex of the kyphosis, the anterior elements of the mid-thoracic spine are subjected to high compressive loads (White et al 1977). Progressive wedge deformity of the vertebral bodies and disc space nar­ rowing are common, even in relatively young individuals (Wood et al1995). These anatomical changes can reduce the mobility of the mid-thoracic motion segments and ribs, par­ ticularly in axial rotation and extension. This pattern of movement restriction is commonly seen in patients with chronic postural pain associated with sustained loading activities. In older patients, mid-thoracic mobility may be further reduced due to the preferential development of anterior vertebral osteophytes in this region (Nathan 1962). On physical examination, a region of relatively limited mid­ thoracic motion may be observed during trunk rotation, which is more evident when rotation is performed with the

arms elevated (Fig. 6.2). This is often associated with com­ pensatory contralateral lateral flexion of the lumbar spine and cramp-like discomfort in the lower thorax due to the increased torsional loading transferred to this region. In extension, a physical barrier to movement may occur due to the reduced disc height, which would promote early approximation of the bony posterior elements. The influ­ ence of these anatomical changes on mid-thoracic extension should be considered in clinical tests which involve passive physiological movement and overpressure. This physical barrier to extension should also be considered when pre­ scribing mobility and posture correction exercises for the thoracic spine. Anatomical variation in the low thoracic spine, particu­ larly the thoracolumbar junction, should be considered in the examination of movement in this region. The transition from a coronal to sagittal zygapophysial joint orientation may be gradual or abrupt resulting in individual differ­ ences in patterns of segmental mobility. The application of motion palpation and mobilization techniques should account for the relatively limited potential for extension and rotation, particularly under weight-bearing conditions.

Figure 6.2 A patient with restricted mid-thoracic rotation demonstrates reduced movement of the mid-thoracic region on movement testing (A). This limitation of movement is more evident when tested in relative thoracic extension (arms elevated) (8), and is associated with compensatory contralateral flexion of the lumbar spine.

Clinical biomechanics of the thoracic spine including the ribcage


Manipulative techniques applied to this region which involve end-range extension or rotation have the potential to produce discomfort or even injury to the joint surfaces or related peri-articular tissues (Singer & Giles 1990). Accessory motion palpation techniques have been advo­ cated for the assessment of range and quality of segmental motion in patients with thoracic spine pain (Magarey 1994). In particular, changes in the through-range resistance to movement (stiffness) in response to posteroanterior (PA) forces applied to the spinous processes may assist in the identification of a symptomatic segment. In asymptomatic subjects, the PA stiffness of the thoracic vertebral segments increases from an average of 9.1 N/mm at T4 to 11.4 N/mm at no (Edmondston et al 1999). Departure from this seg­ mental increase in PA stiffness may be indicative of abnor­ mal motion segment function if associated with a relevant symptom response. The thoracic spine is supported by the compressible ribcage such that assessment of PA stiffness may be strongly influenced by ribcage stiffness. However, ribcage stiffness, measured via sternal loading, is signifi­ cantly lower than the PA stiffness of the thoracic spine and accounts for only 33% of the variation between individuals (Edmondston et al 1999). This suggests that factors other than ribcage stiffness determine the movement response to PA motion palpation tests in the thoracic spine. Posteroanterior load applied to the thoracic spine there­ fore results in a global movement of the spine and ribcage and a more specific movement of the loaded segment. One possible influence on the response to PA loading in the tho­ racic spine is the orientation of the applied force. The appli­ cation of PA force to the spinous process induces anterior translation and posterior rotation (extension) of the related vertebral segment. When a movement force of 200 N is directed anteriorly or perpendicular to the spinal curvature,

a resultant anterior translation of equivalent force is accom­ panied by an extension moment of up to 5.5 Nm (Lee 1989). In contrast, an equivalent force directed towards the verte­ bral body eliminates the extension moment but induces a longitudinal force of up to half the applied load (Lee 1989). Therefore, the movement response to PA accessory motion palpation in the thoracic spine may be influenced by the method in which the test is applied. Consistency in the method of application is required to achieve comparable responses on subsequent testing occasions.

The thoracic spine and ribcage complex has been a rela­ tively limited focus for biomechanical research. This can be attributed to the complex interaction between the spine and ribcage during movement, and technical difficulties, which limit the potential for direct measurement of vertebral and rib motion. Despite this, a better understanding of the bio­ mechanics of the thoracic spine is beginning to emerge. This review provides a summary of the response to load­ bearing and the adaptations to the dual requirement for stability and mobility. Regional variations in thoracic spine kinematics reflect the influence of the anatomical diversity of this region of the spine, and recognition of this is impor­ tant in the application and interpretation of clinical tests and treatment techniques in manual therapy practice.

thoracic spine ribcage biomechanics coupled motion

Anderson R 1992 The back pain of bus drivers. Spine 17: 1481-1488 Andriacchi T, Schultz A, Belytschko T, Galante J 1974 A model for studies of mechanical interactions between the human spine and rib cage. Journal of Biomechanics 7: 497-507 Berg E E 1993 The sternal-rib complex: a possible fourth column in thoracic spinal fractures. Spine 18:1916-1919 Boden S 0, Wiesel S W 1990 Lumbosacral segmental motion in normal individuals. Have we been measuring instability properly? Spine 15: 751-757 Bogduk N 1986 The anatomy and function of the lumbar back muscles. In: Grieve G P (ed) Modem Manual T herapy. Churchill Livingstone, Edinburgh, Ch 13, pp 138-145 Bojadsen T W, Silva E S, Rodrigues A J, Amadio A C 2000 Comparative study of Mm. multifidi in lumbar and thoracic spine. Journal of Electromyography and Kinesiology 10: 143-149 Boyle J J W 1999 Is the pain and dysfunction of shoulder impingement lesion really second rib syndrome in disguise? Two case reports. Manual Therapy 4:44-48 Culham E, Peat M 1993 Functional anatomy of the shoulder complex. Journal of Orthopaedic and Sports PhYSical Therapy 18:342-50 Daggfeldt K, T horstensson A 1997 The role of intra-abdominal pressure in spinal loading. Journal of Biomechanics 30: 1149-1155 Davis P R 1959 The medial inclination of the human thoracic intervertebral articular facets. Journal of Anatomy 93:68-74 DeTroyer A, Estenne M 1988 Functional anatomy of the respiratory muscles. In: Belman M J (ed) Respiratory muscles: function in health and disease. Saunders, Philadelphia Edmondston S J, Singer K P, Day R E, Breidahl PO, Price R I 1994a In­ vitro relationships between vertebral body density, size and compressive strength in the elderly thoracolumbar spine. Clinical Biomechanics 9: 180-186 Edmondston S J, Breidahl W H, Singer K P, Day R E, Price R I 1994b Segmental trends in cancellous bone structure in the thoracolumbar spine: histological and radiological comparisons. Australasian Radiology 38: 272-277 Edrnondston S J, Allison G T, Althorpe B M, McConnell D R, Samuel K K 1999 Comparison of ribcage and posteroanterior thoracic spine stiffness: an investigation of the normal response. Manual Therapy 4:157-162 Edmondston S J, Allison G T, Dahl B-R, Look D, Poirier D, Wapnah M 2000 Trunk muscle and postural responses to incremented spinal loading. International Federation of Orthopaedic Manipulative Therapists 7th Scientific Conference, Perth, Australia Evjenth 0, Hamberg J 1984 Muscle stretching in manual therapy: a clinical manual. Alfta Rehab, Sweden



Feiertag M A, Horton W C, Norman J T, Proctor F C, Hutton W C 1995 The effect of different surgical releases on thoracic spinal motion. Spine 20: 1604-1611 Gibbons P, Tehan P 1998 Muscle energy concepts and coupled motion of the spine. Manual Therapy 3: 95-101 Goh S, Price R I, Leedman P J, Singer K P 1999 The relative influence of vertebral body and intervertebral disk shape on the thoracic kyphosis. Clinical Biomechanics 14: 439-448 Goh S, Price R I, Leedrnan P J, Singer K P 2000 A comparison of three methods for measuring thoracic kyphoSiS: implications for clinical studies. Rheumatology 39: 310-315 Gregersen G G, Lucas D B 1967 An in vivo study of the axial rotation of the human thoracolumbar spine. Journal of Bone and Joint Surgery 49A: 247-262 Harris J D, Holmes T G 1996 Ribcage and thoracic spine. Physical Medicine and Rehabilitation Clinics of North America 7: 761-771 Harrison D E, Harrison D D, Troyanovick S J 1 998 Three-dimensional spinal coupling mechanics. I: A review of the literature. Journal of Manipulative and Physiological Therapeutics 21: 101-113 Hinkley H J, Drysdale I P 1995 Audit of 1000 patients attending the clinic of the British College of Naturopathy and Osteopathy. British Osteopathic Journal 16: 1 7-22 Hodges P W, Gandevia S C 2000 Changes in intra-abdominal pressure during postural and respiratory activation of the human diaphragm. Journal of Applied Physiology 89: 967-976 Horst M, Brinckmann P 1981 Measurement of the distribution of axial stress on the end-plate of the vertebral body. Spine 6: 217-232 Jiang H, Raso J V, Moreau M J, Russell G, Hill D L, Bagnall K M 1994 Quantitative morphology of the lateral ligarnents of the spine: assessment of their importance in maintaining lateral stability. Spine 19: 2676-2982 Klausen K 1965 The form and function of the loaded human spine. Acta Physiologica Scandinavica 65: 1 76-190 Koeller W, Meier W, Hartmann M 1984 Biomechanical properties of the human intervertebral discs subjected to axial compression: a comparison of lumbar and thoracic discs. Spine 9: 725-733 Lee M 1989 Mechanics of spinal joint manipulation in the thoracic and lumbar spine: a theoretical study of posteroanterior force techniques. Clinical Biomechanics 4: 249-251 Lee D 1993 Biomechanics of the thorax: a clinical model of in vivo function. Journal of Manual and Manipulative Therapy 1 : 1 3-21 Lee D 1994 Manual therapy for the thorax: DOPC, British Columbia Leong J C Y, Lu W W, Luk K K D, Karlberg E M 1999 Kinematics of the chest cage and spine during breathing in healthy individuals and in patients with adolescent idiopathic scoliosis. Spine 24: 1310-1315 Lindgren K A, Leino E 1988 Subluxation of the first rib: a possible thoracic outlet syndrome mechanism. Archives of Physical Medicine and Rehabilitation 69: 692-695 Linton S J, Hellsing A-L, Hallden K 1998 A population-based study of spinal pain among 35-45-year-old individuals. Spine 23: 1457-1463 Loring S H, Woodbridge J A 1991 Intercostal muscle action inferred from finite-element analysis. Journal of Applied Physiology 70: 2712-2718 Lovett R W 1905 The mechanism of the normal spine and its relation to scoliosis. Boston Medical and Surgical Journal 13: 349-358 Lowcock J 1991 Thoracic joint stability and clinical stress tests. Orthopaedic Division of the Canadian Physiotherapy Association Newsletter (Nov IDee) Macintosh J E, Bogduk N 1994 The anatomy and function of the lumbar back muscles. In: Boyling J D, Palastanga N (eds) Grieve's Modem Manual Therapy. Churchill Livingstone, Edinburgh, pp 189-209 Magarey M E 1994 Examination of the cervical and thoracic spine. In: Grant R (ed) Physical therapy for the cervical and thoracic spine, 2nd edn. Churchill Livingstone, Edinburgh

Malmivaara A, Videman T, Kuosma E, Troup J D G 1987 Facet joint orientation, facet and costovertebral joint osteoarthritis, disc degeneration, vertebral body osteophytosis, and Schmorl's nodes in the thoracolumbar junctional region of cadaveric spines. Spine 12: 458-463 Manns R A, Haddaway M J, McCall I W, Pullicino V C, Davie M W J 1996 The relative contribution of disc and vertebral morphometry to the angle of kyphosis in asymptomatic subjects. Clinical Radiology 51 : 258-262 Martinez J B, Oloyede V 0, Broom N D 1997 Biomechanics of load­ bearing of the intervertebral disc: an experimental and finite element modeL Medical Engineering and Physics 19: 145-156 Mitchell F L, Mitchell P K G 1995 The muscle energy manuaL MET Press, East Lansing, Michigan Morris J M, Lucas B D, Bresler B 1961 Role of the trunk in stability of the spine. Journal of Bone and Joint Surgery 43A: 327-351 Nathan H 1962 Osteophytes of the vertebral column. Journal of Bone and Joint Surgery 44A: 243-268 Occhipiniti E, Colombini D, Grieco A 1993 Study of distribution and characteristics of spinal disorders using a validated questionnaire in a group of male subjects not exposed to occupational spinal risk factors. Spine 18: 1150-1159 Oda I, Aburni K, Duosai L, Shono Y, Kaneda K 1996 Biomechanical role of the posterior elements, costovertebral joints, and rib cage in the stability of the thoracic spine. Spine 21: 1423-1429 Ortengren R, Andersson G B J 1977 Electromyographic studies of trunk muscles with special reference to the functional anatomy of the lumbar spine. Spine 2: 44-52 O'Sullivan PB 2000 Lumbar segmental 'instability': clinical presentation and specific stabilizing exercise management. Manual Therapy 5: 2-12 Pal G P, Routal R V 1986 A study of weight transmission through the cervical and upper thoracic regions of the vertebral column in man. Journal of Anatomy 148: 245-261 Pal G P, Routal R V 1987 Transmission of weight through the lower thoracic and lumbar regions of the vertebral column in man. Journal of Anatomy 152: 93-105 Panjabi M M 1992 The stabilising system of the spine. II: Neutral zone and instability hypothesis. Journal of Spinal Disorders 5: 390-397 Panjabi M M, Goel V K 1982 Physiologic strains in the lumbar spinal ligaments. Spine 7: 192-203 Panjabi M M, Brand R A, White A A 1976 Mechanical properties of the human thoracic spine. Journal of Bone and Joint Surgery 58A: 642-652 Panjabi M M, Hausfeld J N, White A A 1981 A biomechanical study of the ligamentous stability of the thoracic spine in man. Acta Orthopaedica Scandinavica 52: 315-326 Panjabi M M, Krag M H, Dimnet J C, Walter S D, Brand R A 1984 Thoracic spine centres of rotation in the sagittal plane. Journal of Orthopaedic Research 1 : 387-394 Panjabi M M, Aburni K, Daranceau J 1989 Spinal stability and intersegmental muscle forces. Spine 14: 194-200 Pearsall D J, Reid J G 1992 Line of gravity relative to the upright vertebral posture. Clinical Biomechanics 7: 80-86 Pooni J S, Hukins D W, Harris P F, Hilton R C, Davies K E 1986 Comparison of the structure of human intervertebral discs in the cervical, thoracic and lumbar regions of the spine. Surgical and Radiological Anatomy 8: 1 75-182 Putz V R, Muller-Gerbl M 2000 Ligaments of the human vertebral column. In: Giles L G F, Singer K P (eds) Clinical anatomy and management of thoracic spine pain. Butterworth Heinemann, Oxford Rickenbacher J, Landolt A M, Theiler K 1985 Applied anatomy .of the back. Springer-Verlag, Berlin Saumarez R C 1986 An analysis of possible movements of the upper rib cage. Journal of Applied Physiology 60: 678-689

Clinical biomechanics of the thoracic spine including the ribcage


Scott J E, Bosworth T R, Cribb A M, Taylor J R 1994 The chemical morphology of age-related changes in human intervertebral disc glycosaminoglycans from cervical, thoracic and lumbar nucleus pulposus and annulus fibrosis. Journal of Anatomy 184: 73-82 Shea K G, Schlegel J D, Bachus K N, Dunn H K, West J R 1996 The contribution of the rib cage to thoracic spine stability. In: Proceedings of the International Society for the Study of the Lumbar Spine, Vermont Singer K P 1997 Pathomechanics of the aging thoracic spine. In: Lawrence D (ed) Advances in chiropractic. Mosby Yearbook, Chicago Singer K P, Edmondston S J 2000 The enigma of the thoracic spine. In: Giles L G F, Singer K P (eds) Clinical anatomy and management of thoracic spine pain. Butterworth Heinemann, Oxford Singer K P, Giles L G F 1990 Manual therapy considerations at the thoracolumbar junction: an anatomical and functional perspective. Journal of Manipulative and Physiological Therapeutics 13: 83-88 Singer K P, Malmivaara A 2000 Pathoanatomical characteristics of the thoracolumbar junctional region. In: Giles L G F, Singer K P (eds) Clinical anatomy and management of thoracic spine pain. Butterworth Heinemann, Oxford Singer K P, Day R E, Breidahl P D 1989 In vivo axial rotation at the thoracolumbar junction: an investigation using low dose CT in healthy male volunteers. Clinical Biomechanics 4: 145-150 Singer K P, Edmondston S J, Day R E, Breidahl W H 1994 Computer­ assisted and Cobb angle determination of the thoracic kyphosis: an in-vivo and in-vitro comparison. Spine 19: 1381-1384 Singer K P, Edmondston S J, Day R E, Breidahl P D, Price R I 1995 Prediction of thoracic and lumbar vertebral body compressive strength: correlations with bone mineral density and vertebral region. Bone 17: 167-174 Sobel J S, Kremert I, Winters J C, Arendzen J H, de Jong B M 1996 The influence of the mobility in the cervicothoracic spine and the upper

ribs (shoulder girdle) on the mobility of the scapulohumeral joint. Journal of Manipulative and Physiological Therapeutics 19: 469-474 Stokes I A F 2000 Biomechanics of the thoracic spine and ribcage. In: Giles L G F, Singer K P (eds) Clinical anatomy and management of thoracic spine pain. Butterworth Heinemann, Oxford Toppenberg R M, Bullock M I 1986 The interrelationship of spinal curves, pelvic tilt and muscle lengths in the adolescent female. Australian Journal of PhYSiotherapy 32: 6-12 Veldhuizen A G, Scholten P J M 1987 Kinematics of the scoliotic spine as related to the normal spine. Spine 12: 852-858 Walker M L, Rothstein J M, Finucane S D, Lamb R L 1987 Relationships between lumbar lordosis, pelvic tilt, and abdominal muscle performance. Physical Therapy 67: 512-516 White A A 1969 An analysis of the mechanics of the thoracic spine in man. Acta Orthopaedica Scandinavica 127(Suppl.): 8-92 White S G, Sahrmann S A 1994 A movement system balance approach to management of musculoskeletal pain. In: Grant R (ed) Physical therapy for the cervical and thoracic spine, 2nd edn. Churchill Livingstone, Edinburgh White A A, Panjabi M M, Thomas C L 1977 The clinical biomechanics of kyphotic deformities. Clinical Orthopaedics and Related Research 128: 8-17 Wilke H-J, Wolf S, Claes L E, Arand M, Wiesend A 1995 Stability increase of the lumbar spine with different muscle groups. Spine 20: 192-198 Willems J M, Jull G A, Ng J K-F 1 996 An in-vivo study of the primary and coupled rotations of the thoracic spine. Clinical Biomechanics 11: 311-316 Wisleder D, Smith M B, Mosher T J, Zatsiorsky V 2001 Lumbar spine mechanical response to axial compression load in vivo. Spine 26(18): E403-409 Wood K B, Garvey T A, Gundry C, Heithoff K B 1995 Magnetic resonance imaging of the thoracic spine. Journal of Bone and Joint Surgery 77A: 1631-1638



Chapter 7

Clinical biomechanics of the ' lumbar spine
J. Cholewicki. s. P. Silfies

67 68 71 71 71 71 72

Theoretical basis of structural analyses of equilibrium and stability Estim.,ting spine loads Optimization methods EMG-assisted methods

Stability of the lumbar spine

Trunk muscles as variable stiffness springs Spine stabilizing role of trunk muscles stability
73 74 75

Role of intra-abdominal pressure in spine Role of abdominal belts and lumbar supports in spine stability Biomechanics of spine injury and pain injury injury
76 77 77

Equilibrium based concept of musculoskeletal Stability based concept of musculoskeletal Explanation for injury occurrence under very low loads pain
78 78 79 80

Muscle recruitment patterns and low back Motor control of spine stability and low back pain Cause or effect?

Impairment or adaptation? control
81 81

Clinical relevance of trunk stability and motor Clinical assessment of trunk stability and motor control Implications for rehabilitation strategies

There are three basic mechanical functions for the lumbar spine: protection of the spinal cord and nerve roots, per­ mitting motion between the pelvis and thorax, and trans­ mission of loads between the pelvis and thorax. Failure in any one of these three mechanical functions could result immediately in, or lead to, a clinical problem. The topic of spinal kinematics has been covered in a number of biome­ chanics texts and the discussion of spinal cord and nerve root protection is probably better suited by an anatomical approach. However, the biomechanics of spinal load trans­ mission in the context of mechanical equilibrium, stability and injury mechanisms has considerable implications for clinical evaluation and treatment strategies and it will be the focus of this chapter. The support of loads that arise from interaction between external and muscular forces is probably the single most important mechanical function of the spine. Because the muscles act through a relatively small moment arm in rela­ tion to the moment arm of external forces, the spine sus­ tains extremely high loads. Not surprisingly, mechanical factors are often identified as the primary cause in a large percentage of low back disorders (Cherkin et al 1992, Deyo & Weinstein 2001, Kerr et al 2001, Marras et al 1995, McCowin et aI1991). While other psychosocial and patho­ physiological factors leading to low back pain (LBP) have also been identified, this chapter will focus solely on the mechanical factors. Therefore, when referring to LBP or injury throughout this chapter we are implicitly consider­ ing only the mechanical causes. Currently, the assessment of spine loads and subse­ quently the elucidation of the mechanisms of injury are possible only through biomechanical modelling. Other methods of in vivo load measurement exist, such as instru­ mented implants (Rohlmann et al 2000), but they are very limited owing to their invasiveness, patient population and technological constraints. Therefore, much of this chapter is devoted to the discussion of biomechanical equilibrium and stability models and conceptual models of lumbar



spine injury. We will summarize the current research and discuss the application of an instability/motor control injury model to the clinical evaluation and treatment of patients with mechanical low back dysfunction.

For the safe support of loads by any mechanical structure, its material must withstand the load and the structure itself must be stable. This leads to a two-step approach in the structural analysis of mechanical systems. The first level analysis relies on the force and moment equilibrium condi­ tions for the computation of loads arising at various loca­ tions of interest in the structure. Depending on the system, this analysis can be static or dynamic. In the latter case, the inertial forces are included in the equations of equilibrium. The standard approach is to draw a free body diagram, which is a representation of an isolated part of a system with all of the forces and moments acting on it. For exam­ ple, to estimate the loads acting at the L3-4 intervertebral joint during lifting, a free body diagram is drawn (Fig. 7.1). The sum of forces and moments arising from the upper body mass, muscle action, weight held in hands and the joint reaction forces must be zero to satisfy the static equi­ librium condition. The unknown muscle and joint reaction

Figure 7.1 A free body diagram of the lumbar spine for the ca lcu­ lation of the reaction forces (R) acting on the L3-4 intervertebra l joint. Because the moment arm (rL) of the load (L) is usua l ly much greater than the moment arm (rM) of the combined erector spine muscles, their force (M) must be much greater to ba lance the moment equi librium e quation. From the force equi librium equation, it fo l lows that the joint reaction force is the sum of load and mus­ c le force (R L + M).

forces can be computed by solving the moment and force equations simultaneously. One should note that the large resultant joint compression force stems mainly from the muscle action and can be several times greater than the com­ bined upper body weight and the weight held in hands. The second level analysis examines whether the equilib­ rium state defined in the first level analysis is stable. The terms 'stability' and 'instability' referring to given joints or systems of joints are often misused in biomechanics litera­ ture. Within spine biomechanics, the stability concept is complicated by several clinical definitions of segmental instability consisting of a variety of diagnostic findings (White & Panjabi 1990). Several attempts to clarify and stan­ dardize the terms 'stability' and 'instability' have been made. Pope & Panjabi (1985) proposed that 'stability' (or 'instabil­ ity') is a mechanical entity and should be treated as such. Definitions should not be based on suspected injury mecha­ nism or 'clinical history'. Similarly, Ashton-Miller & Schultz (1991) called for a standard use of these terms in biomechan­ ics. However, both 'clinical instability' and 'mechanical sta­ bility/instability' may be used concurrently if a clear understanding of the distinctions between them exists. From a mechanical point of view, stability analysis refers to the study of an unperturbed state of a system. A perturbation is applied and certain quantities, which characterize the state of the system at any time, are measured. If, as the system goes from the unperturbed to perturbed state, the changes in those quantities do not exceed their earlier established measures, the unperturbed state is called stable. If these quantities exceed their earlier established measures, the unperturbed state is unstable (Leipholz 1987). An example of a clinical application of this definition is testing of patients' static standing or seated balance. A clinician provides perturbation to a patient to ascertain his ability to maintain balance and to return to equilibrium or a stable state. If the patient fails to maintain balance or his sway exceeds some normative dis­ tance, his stance will be classified as unstable. The state of a dynamic system is generally characterized with parameters describing its motion. Therefore, the sta­ bility of the dynamic system will refer to the stability of its unperturbed motion. A control mechanism(s), if present, becomes an integral part of such a system and will also affect its stability. For example, a constant velocity and intended trajectory can describe unperturbed motion of a car on cruise control. A multitude of system parameters will affect this car's stability when it encounters a perturbation such as a bump on a road or a gust of wind: stiffness of the suspension, friction between the tyres and the road or the quality of the cruise control, to mention only a few. Similarly, in the most general terms, spine stability refers to the capability of maintaining and controlling physiological spine movements and it includes a motor control system. Hypermobility, for example, is one of the spine characteris­ tics. It does not necessarily imply instability of the entire spine system, especially if it can be adequately compensated for and controlled resulting in coordinated and pnysiologi­ cal spine movement. Unfortunately, current biomechanical

C l in i ca l b i o mechan i cs of t he l u m bar sp ine


models are still limited to static analyses of stability, although the mathematical theory is available to study fully dynamic systems (Leipholz 1987). These models focus on muscle and joint stiffness and various muscle recruitment patterns. However, some inferences about motor control and the dynamic stability of the spine can be made by com­ paring static spine stability obtained from these models and patients' responses to various perturbations (see p. 78) . For example, the dynamic response of a patient to sudden trunk loading depends on the static stability of the spine exhibited prior to sudden loading and the muscle reflex response (motor control) after sudden loading (Cholewicki et al2000). In a static example, let us examine stability of the equilib­ rium states of the four mechanical systems represented by balls on different surfaces in Figure 7.2. Each system is in a static equilibrium. Upon perturbation, only the balls in the last two examples will return to their original equilibrium positions. These two systems are therefore stable. The balls in the first two examples will be displaced away from their original equilibrium positions following the perturbations, indicating unstable equilibrium states of these systems. The mathematical formulation of the stability problem in elastostatic systems such as one considered above relies on the minimum potential energy principle. A mechanical system is stable only if its total potential energy is at a relative mini-

mum. In other words, any mechanical perturbation would cause the potential energy of a stable system to rise and it would then tend to return to its relative equilibrium. It can easily be seen in Figure 7. 2 that the potential energies of stable systems are at their respective minima. It should also be noted that static equilibrium is a necessary but not a sufficient con­ dition for stability. If a system is not in equilibrium, it is not stable by definition. Furthermore, the stability state can be quantified with the measure of the curvature of the potential energy. The larger the curvature (depth) of the potential energy in the vicinity of its minimum, the more stable the sys­ tem is. For example, the system represented in Figure 7.20 is more stable than the system represented in Figure 7.2C. In a more realistic example of an inverted pendulum resembling a spine model, the change in potential energy in various forms must be considered (Fig. 7.3). The total potential energy (V) of such a system after the perturbation is the difference between the elastic energy (U) stored in springs and the work (W) performed by the external load: V=U-W (equation 1)

Furthermore, the elastic energy stored in springs is pro­ portional to their stiffness (k) (equation 2)


where Xl and x2 are the changes in the springs' length. The work performed by the external load (L) is given by: W=L e (equation 3)







F igure 7.2 A simple mechanica l system i l lust rating the p rincip le of the mini mum potentia l energy. In a l l f our cases (A, B, C, and D) the syste m satisfies static equi libriu m c onditi ons . However, on ly the C and D cases are stab le, because each of these systems' p otentia l energy is at its respective mini mu m.

Figure 7.3 A simp lified spine model i l lustrating the energy app roach to ana lysis of stabi lity. The total potentia l energy of such a system after the perturbati on is the p otent ia l energy stored in springs ( musc les) minus the w ork perf or med by the externa l load (L). Stiffer springs (k) store more p otentia l energy and create a more stable system.



Now it remains to examine the total potential energy for its behaviour around the equilibrium state. Mathematically, the first derivative of the potential energy must be equal to zero to satisfy the static equilibrium requirement and the second derivative must be greater than zero for stability (equations 4 and 5). The second derivative also quantifies the curvature of the potential energy (V" > 0 implies the concave surface) and hence the stability of the system. V'

0 static equilibrium

(equation 4) (equation 5)

V" > 0 stability

Equations 1 and 5 can be interpreted in the following way. If upon the perturbation, the amount of stored elastic energy is greater than the work performed by the external forces, the overall energy of the system will rise. Such a sys­ tem is stable and it will return to its original equilibrium configuration. In contrast, if the elastic potential energy stored in springs is less than the external work, the system is unstable and it will continue to deform seeking the min­ imum potential energy - it will buckle. It can be seen by combining equations 1, 2 and 3 that the stiffer the springs are, the more stable the system is. This is because more elas­ tic potential energy is stored upon the perturbation. Similarly, the larger the external load is, the less stable the system. The final observation is that an elastostatic system, or the forces acting upon it, need not be symmetrical for it to be stable, as long as the static equilibrium is satisfied. The minimum potential energy principle is one classical approach used to determine the stability criteria of an elas­ tic system with multiple degrees of freedom (Fig. 7.4). The

only difference from the previous examples is that now the potential energy forms a multidimensional surface around the equilibrium state (number of dimensions equals the number of degrees of freedom). Therefore, partial, second order derivatives of the potential energy with respect to each coordinate must now be greater than zero to satisfy stability criteria. In other words, the potential energy sur­ face must be concave in every direction at the point of equi­ librium to form the minimum and for the entire system to be stable. If this surface is convex in the direction of any one degree of freedom, the entire system will be unstable. The average curvature of the potential energy surface - termed the stability index (SI) by Cholewicki & McGill (1996) - can be used to quantify the relative stability of a multi-degree of freedom system. In the lumbar spine, muscles, along with ligaments and other passive tissues, play the stabilizing role by momen­ tarily storing the elastic potential energy in response to mechanical perturbations. The muscles act as variable stiffness springs whose stiffness is proportional to the muscle force. If the spine is sufficiently stable it will resist external perturbations without the need for active feed­ back control. In other words, the spine will return to its equilibrium state after the perturbation even if no change in muscle activation had occurred. Due to inherent delays in feedback loops, active control of relatively small and transient perturbations may not be efficient and/or effec­ tive. Several issues pertaining to the stability of the lum­ bar spine will be discussed in more detail later in the chapter. In summary, a complete biomechanical assessment of spine injury potential, injury mechanisms or the biome­ chanical evaluation of the effectiveness of various preven­ tion and rehabilitation approaches should encompass the two analytical steps outlined above. The estimation of tis­ sue loads is necessary to assess the risk of tissue failure under various spine-loading scenarios. However, tissue integrity alone does not assure the structural stability of the spine. Therefore, the assessment of spine stability is also necessary to further elucidate the potential or effects of structural failure due to buckling.

Estimating spine loads

Figure 7.4 A schematic of a mu ltidegree-of-f reedo m spine model. If one of the deg rees of f reedom beco mes unstable, the enti re st ructure is unstable and it wi l l buck le under the load (L). Muscle and liga ments must p rovide stabi lity with a coo rd inated muscle rec ruit ment pattern.

The biomechanical analysis of spinal loads begins with a free body diagram and the equations for static or dynamic equilibrium of forces and moments. Models containing even minimal anatomical detail result in a mathematically indeterminate problem caused by existence of multiple tis­ sues that can generate or support forces and moments about a given joint. There are two basic methods for s?lving the problem of mathematical indeterminacy in a biome­ chanica1 spine model: optimization and EMG-assisted approaches. Each of these methods offers a number of distinctive assets and liabilities.

Clinical biomechanics of the lumbar spine


Optimization methods

optimization method relies on formulating an objective function that serves as a criterion for selecting a unique solution of force partitioning among various tissues out of the infinitely large set of viable solutions. This criterion may consist of minimizing the sum of muscle forces (Yettram & Jackman 1980), the sum of muscle stress, disc compression, joint shear force, or some combination of these (e.g. Bean et al 1988, Schultz et al 1983). Because the optimization solution converges on a singular set of muscle forces to meet the moment constraints, it is insensitive to the transient changes in load sharing among agonist mus­ cles during the exertion. Current objective functions are not able to respond to the many different ways in which mus­ cles are recruited to perform similar tasks even when the kinematics or resultant moment patterns are the same. A popular objective function in many low back optimization models, minimization of muscle stress and disc compres­ sion, predicts no antagonist muscle co-activation (co-con­ traction), defined as the contraction of muscles above the minimum stress necessary to satisfy the moment equilib­ rium about a given joint (Hughes et al 2001). In turn, this optimization scheme underestimates the joint compression forces during isometric exertions by 23-43% when com­ pared with an EMG-assisted approach (Cholewicki et al 1995, Hughes et al 1995). The antagonistic co-activation of trunk muscles is often demonstrated with EMG during many activities (Granata & Orishimo 2001, Lavender et al 1993, 1992b). Among other hypotheses, the antagonist mus­ cle co-activation is explained as necessary for providing mechanical stability to the spinal column (Bergmark 1989, Cholewicki & McGill 1996, Crisco & Panjabi 1991, Gardner­ Morse et al 1995, Granata & Marras 2000).

profiles. Minimal adjustments are applied to the individual muscle forces estimated initially from EMG, to balance all moment and force equations. The EMGAO combines some principal advantages of the optimization and EMG-assisted methods. It preserves the physiologically observed (through EMG) muscle activation patterns while satisfying the equilibrium constraint equations exactly (Cholewicki et al 1995). Despite the obvious advantage of better physiological accuracy of the EMG and EMGAO spine models, they require complex data acquisition and processing method­ ologies. For the applications that require only rough esti­ mates of spinal loading, optimization or even single muscle equivalent models may suffice (Kingma et al 1998, McGill et al 1996, van Dieen & de Looze 1999b). However, simula­ tions with such models will always produce identical results for the same loading (input) conditions. It is not pos­ sible to detect differences in neuromuscular control between the subjects or the different features among the 'normal' and 'abnormal' muscle activation patterns or their effects on spine forces. The EMG-assisted models are better suited for this purpose because their input is biologically sensitive to the various patterns of muscle recruitment.
Stability of the lumbar spine

EMG-assisted methods

An EMG-assisted method partitions the forces among the muscles according to their normalized EMG activity, cross­ sectional areas and assumptions regarding their maximum force-generating potential (Granata & Marras 1995, McGill 1992a). In the dynamic version of this method, predicted muscle forces are further modulated with coefficients accounting for instantaneous muscle length, velocity of contraction and passive elastic contributions. EMG­ assisted partitioning of muscle forces is inherently consis­ tent with physiologically observed muscle activation patterns. However, due to imperfections in EMG recording and processing and anatomical modelling, the simultane­ ous moment equations in three dimensions are not satis­ fied very well in complex tasks (Granata & Marras 1995, McGill 1992a). To remedy the equilibrium problem, a hybrid approach, termed EMG-assisted optimization (EMGAO), was devel­ oped (Cholewicki & McGill 1994). In this method, an opti­ mization algorithm is used to satisfy the equilibrium equations in a way that provides the best possible match between the predicted muscle forces and their myoelectric

In vitro estimates of the critical loads of isolated osteoliga­ mentous spine segments highlight the importance of the mechanical stability of the spine. In a classic experiment Lucas & Bresler (1961) determined the critical load for a thoracolumbar spine to be approximately 20 N (4.5 lb). This indicates that the spine is unable to sustain compressive loads and will buckle under very low loads. A later replica­ tion of this study established the critical load for a lumbar spine to be approximately 90 N (20 lb) (Crisco et al 1992). The lumbar spine must support an upper body weight four to five times greater than its buckling threshold load. If any additional external forces are acting on the torso, spine sta­ bility surfaces as the most important issue in supporting and transmitting such loads. It becomes clear that the static or dynamic equilibrium analysis in a spine model is not enough to study the above phenomena. It is now necessary to incorporate structural stability analysis tools into the bio­ mechanical models.
Trunk muscles as variable stiffness springs

Stability analysis has been applied to spine modelling only relatively recently (Bergmark 1989, Cholewicki & McGill 1996, Crisco & Panjabi 1991, Gardner-Morse et al 1995, Granata & Marras 2000). To our knowledge, Bergmark (1989) was the first to incorporate a spring-like short-range muscle stiffness into the calculations of stability in a multi­ ple degrees of freedom spine model. Short-range muscle stiffness, also called high frequency stiffness, relates small changes in the muscle length and force, such that they will not result in the change of cross-bridge attachment



distribution. The mechanical properties of the whole mus­ tic (conservative) (Rack & Westbury 1973, 1974) and can be modelled with a mechanical spring (Hogan 1990). The short-range stiffness of the muscle has been shown to be linearly related to the muscle force (Morgan 1977, Zahalak cle/tendon unit within this short range are essentially elas­ greater spine compression force penalty. In fact, a low level of trunk muscle co-contraction, in the range of 1-2% of a maximum voluntary exertion, is necessary to maintain the spine in a stable equilibrium around its neutral posture (Cholewicki et al 1997). It is easy to see from the earlier discussion that increased muscle force increases muscle stiffness, which causes more of the elastic potential energy to be stored in the muscles upon the transient perturbations, which in turn leads to greater spine stability. Therefore, there appears to be an ample stability safety margin in tasks that require a lot of muscular effort (Cholewicki & McGill 1996). In contrast, tasks that demand very little muscle activity, such as upright standing with no load, are characterized by low spine stability. The tasks that challenge spine stability the most are those in which spine posture is maintained within its neutral zone, where ligaments are relatively slack, and (equation 6) there are very few muscles activated to stabilize it. It seems reasonable for the neuromuscular system to maintain the most stability during heavy lifting or other high intensity exertion tasks, when the spine buckling would have delete­ rious effects. On the other hand, low energy expenditure may be an objective of the motor control system during standing, sitting or walking tasks that must be sustained over longer periods. Figure 7.5 conceptually compares injury risks due to tissue overload and spine instability as functions of task demand.
In addition to the overall intensity of muscle co-activation,

& Heyman 1979), although some researchers have reported a non-linear relationship (Hatta et al 1988, Stein & Gordon
1986). Beyond this short range the muscle stiffness is mod­ ulated by spinal reflexes and eventually by voluntary Winters et al 1988, Zajac & Winters 1990). responses (Diener et al 1983, Nashner & Cordo 1981, The short-range muscle stiffness (k) can be roughly esti­ inversely proportional to its length (L) (Rack & Westbury 1974): mated as being proportional to the muscle force (F) and


The proportionality constant q varies anywhere between 5 and 100 depending on muscle excitation and tendon-to­ muscle length ratio (Cholewicki & McGill 1995, Crisco & Panjabi 1991). For more accurate estimates of muscle stiff­ ness, especially in dynamic simulations, a distribution­ release and diffusion) (Cholewicki & McGill 1995, 1996, moment model of the muscle activation dynamics (calcium

Zahalak 1986, Zahalak & Ma 1990) or a model with an improved muscle reflex loop (Gielen & Houk 1987, Ramos et al 1990, Stein & Oguztoreli 1984, Winters 1995) is better The ligaments, intervertebral disc and other passive structures also contribute to the stability of the lumbar spine by acting as non-linear springs. Their contribution to spine stability may have been overlooked in the past. The passive stiffness of the osteoligamentous lumbar spine increases significantly with a compressive load placed on the spine. In fact, the osteoligamentous lumbar spine can carry up to 1200 N (270 lb) if this load is distributed to fol­ low the spine curvature (follower load) (Patwardhan et al 1999), which may be the likely in vivo scenario. Nevertheless, more research is necessary to establish the extent of relative sharing of the stabilizing roles between the passive and active (muscles) tissues in the spine. suited. enhanced spring-like muscle performance through an

tecture (Crisco & Panjabi 199 1). Large muscles with greater

the stability of the spinal column depends on muscle archi­ moment arms are more effective in stabilizing the spine than smaller intervertebral muscles. However, each vertebral

body must have at least one muscle fascicle attached and activated, otherwise the spine will always be unstable (Crisco & Panjabi 1991). In addition, for any given activation

Spine stabilizing role of trunk muscles
The effects of different trunk muscle activation patterns on spine stability have been studied through experimentation with stability models of various complexities using both optimization and EMG-assisted methods. Optimization models were shown to be able to predict antagonistic muscle co-activation if the stability criteria were incorpo­ rated into their objective functions (Cholewicki et al 1997, Granata & Marras 2000, Stokes & Gardner-Morse 1999). These studies demonstrated that the antagonistic muscle co-contraction increases spine stability in exchange for a

Figure 7.5 Conceptua l view of the musculoske leta l injury risks as a function of task exe rtion de mand. Like lihood of tissue ove rload and fai lu re inc reases with inc reased task exe rtion. Ho weve r, spine st ructu ra l fa i lu re due to buckling (and in tu rn some t issue ove r­ st raining due to the buckling event) is mo re l ikely to occu r when the musc le fo rces a re low. Adapted f ro m Cholewicki Et McGilr (1996).

Task exertion demand

C l i n i c a l biomech a n ics of the l u m b a r sp i ne


level of the muscles that attach to each lumbar vertebra, there exists an upper limit for the activation of the large muscles that attach only to the pelvis and ribcage (Bergmark 1989). Beyond this limit, the spine becomes unstable. It is analogous to holding a stack of tennis balls by grasping only the top and bottom ones. Each joint must be stabilized prior to acti­ vating large trunk muscles, which apply compressive forces between the ribcage and pelvis ( Fig. 7.6). Based on the above functional dichotomy and on whether the muscles cross a single intervertebral joint or span across all joints from the ribcage to pelvis, Bergmark ( 1989) divided the trunk muscles into 'local' and 'global' systems. The transversus abdominis, portions of the inter­ nal oblique and lumbar multifidus have been labelled as local trunk muscles, whereas the rectus abdominis, external oblique and lumbar erector spinae muscle groups belong to the global muscle system. Unfortunately, the above classifi­ cation and Bergmark's work are often misinterpreted as identifying the muscles that are spine stabilizers and the muscles that are moment generators. While there may be some trunk muscles that are clinically more important than others, this notion is not supported by mechanical stability analyses. All trunk muscles contribute to spine stability and all muscles that cross a given joint contribute to the joint moment. The overall stability of the spine depends on the individual forces, and hence stiffness, of all trunk muscles as well as their relative force magnitudes. The total joint moment is the sum of products of all muscle forces and their respective moment arms. The stability of the lumbar spine is a highly non-linear function of the trunk muscle forces. First, as discussed above, stability depends on both absolute and relative mus­ cle forces. Second, the relative contribution of a muscle to

spinal stability depends on the magnitude and direction of external trunk loading ( Cholewicki & VanV liet 2002). Simulations with muscle 'knock out' in a spine stability model showed that no single muscle group contributed more than 30% to the overall stability of the lumbar spine ( Cholewicki & VanV liet 2002). No single muscle group could be identified as the most important spine stabilizer and no clear distinction was found between the local and global muscles as related to stability. Finally, increased spine stiffness due to spine compression force and the liga­ ment forces that are dependent on spine posture must be also considered among the factors determining the overall stability of the spine.

Role of intra-abdominal pressure in spine stability
Much controversy surrounds the mechanical role of increased intra-abdominal pressure (lAP) in preparation for or during physical exertions. Very high pressures, com­ monly observed during strenuous activities, were origi­ nally hypothesized to reduce the compressive forces on the lumbar spine (Bartelink 1957, Keith 1923, Morris et al1961). The pressure produced within the abdominal cavity exerts a hydrostatic force down on the pelvic floor and up on the diaphragm. This force adds tensile load to the spine and produces trunk extension moment and was therefore assumed to reduce spine compression force. Later, how­ ever, researchers observed that the forceful contraction of abdominal muscles that appears to be necessary to generate IAP would cancel out the tensile force and extensor moment obtained from IAP ( McGill & Norman 1987). In

fact, in vivo intradiscal pressure measurements would sug­ gest that the lumbar spine compression force increases, rather than decreases, with voluntary increase in IAP ( Valsalva manoeuvre) ( Nachemson et al 1986) ( Fig. 7.7). If the transversus abdominis and/or oblique muscles were recruited preferentially to create IAP without the acti­ unloading effect could be achieved with IAP ( Daggfeldt & vation of rectus abdominis, then perhaps a net spinal

Thorstensson 1997, Nachemson et al 1986). Additionally, a small trunk extension moment can be produced with con­ traction of the diaphragm alone ( Hodges et al 2001). The question then arises as to whether people can generate lAP with such a preferential muscle recruitment pattern and without the penalty of additional compressive forces from other longitudinally oriented muscles. Indeed among all minis correlates the best with lAP ( Cresswell & Thorstensson abdominal wall muscles, activation of transversus abdo­

Fig ure 7.6 A sche matic i l lust ration of the re lationship between the mu ltiseg menta l muscles ( muscles that span the pelvis and ribcage), interseg menta l muscles ( muscles that span ind ividual interve rtebra l joints) and sp ine stabi lity. Each inte rve rteb ra l joint must have a musc le fascic le attached and act ivated acco rding to one or both of the two depicted a rchitectu res (C risco Et Panjabi 1991). Fu rthermo re, fo r any g iven activation of the inte rseg menta l muscles, the re exists a li mit for the activation of the multisegmen­ ta l muscles beyond which buckling wi l l occu r.

1989; Cresswell et al 1992, 1994) and it is recruited first in preparation for rapid limb movements ( Hodges & Richardson 1996, 1998, 1999). However, an overall pattern of trunk muscle co-contraction associated with increased IAP was observed by other researchers who hypothesized that it enhances spine stability with a resultant increase in spine compressive load ( Cholewicki et al 1999a, Cresswell McGill & Sharratt 1990). et al 1994, Marras & Mirka 1996, McGill & Norman 1987,



1998, 1999), but they appear to be tightly coupled under steady-state exertions. The concurrent rise in intrathoracic pressure, IAP and muscle co-contraction during physical exertions can easily be explained, based on stability requirements. A high-level physical exertion, such as a lift, throw or jump, requires a rapid contraction of limb and other muscles that originate on the thorax. To execute an effective lift or throw, not only must the 'mechanical slack' be taken up from the muscles prior to the exertion, but a rigid base from which these muscles originate must also be created. Co-contraction of the latissimus dorsi, thoracic erector spinae and intercostal

Abdominal muscles

l t

l t

muscles against the ITP increases the rigidity of the ribcage while the co-contraction of the abdominal wall and lumbar
Back muscles

erector spinae muscles against the lAP increases stability of the lumbar spine (Cholewicki et aI 1999a). Furthermore, ITP helps the contracting diaphragm to increase lAP by reduc­ ing the trans diaphragmatic pressure (Cholewicki et al 2002a). Therefore, the co-contraction of all trunk muscles, including the abdominal wall, erector spinae and latissimus dorsi, along with the increase in IAP and ITP, stiffens both the lumbar spine and the thoracic cage, with a net effect of increased spine compression force. There are other possible mechanical and physiological effects of increased lAP during physical exertion.

Figure 7.7 Int ra-abdo minal p ressu re ( lAP) mechanics. The tensi le force and the t runk extensor mo ment achieved through the action of lAP on the pe lvic f loo r and the diaph rag m is offset or even exceeded by the conco mitant inc rease in the abdominal and back musc le co-cont raction necessa ry to gene rate lAP.
Further support for the spine-stabilizing role of IAP came

Abdominal wall muscles, especially oblique and transver­ sus abdominis, can gain greater mechanical advantage if they contract around the pressurized abdomen than if they collapse inward along their straight lines of action (Cresswell & Thorstensson 1989). A concomitant increase in

the cerebrospinal fluid pressure may act as a safety mecha­ nism by opposing a rise in arterial blood pressure (Porth et aI 1984). Although it has been suggested that exhaling dur­ ing such exertions may reduce blood pressure and mini­ strategy would also reduce lAP, ITP and the level of trunk muscle co-contraction. As a result, reduced spine stability and rigidity of the thorax would compromise the intended physical performance.
In summary, the extremely high IAP levels generated by

untarily generated IAP (Cholewicki et al 1999b). In this study, a significant increase in EMG activity of all major abdominal, lumbar and thoracic muscles was documented when subjects elevated their IAP from its resting value to 40% and 80% of the maximum voluntarily generated pres­ sure. Along with this increase in muscle co-contraction,

from a report of increased trunk stiffness stemming from vol­

mize the risk of a stroke (Narloch & Brandstater 1995), this

trunk stiffness rose significantly by 12% and 32%, respec­
(Cholewicki et al 1999b). Undeniably, this enhancement of

competitive weightlifters (McGill et a1 1990) do not reduce the spine compression force, but rather prevent the collapse of the ribcage and the buckling of the lumbar spine. Likewise, the increased lAP observed in individuals preparing for sudden trunk loading (Cresswell et al 1994) Hemborg & Moritz 1985), likely serves to enhance spine stability prior to any movement. or in patients with non-specific LBP (Fairbank et al 1980,

tively, indicating enhanced stability of the lumbar spine spine stability via IAP and trunk muscle co-contraction came about with the price of increased spine compression force (Gardner-Morse & Stokes 1998, Granata & Marras 2000). Generally, individuals are unable to decouple an increase in lAP from trunk muscle co-contraction during steady­ state exertions (Cholewicki et al 2002a). lAP, intrathoracic pressure (ITP ) and trunk muscle co-contraction are highly correlated regardless of whether subjects attempt to increase IAP without trunk muscle co-contraction or to co­ contract their muscles without elevating their IAP. These entities may dissociate temporarily during transient states for a rapid arm movement (Hodges & Richardson 1996, such as exhaling (Cholewicki et al 2002a) or in preparation

Role of abdominal belts and lumbar supports in spine stability
The notion of the beneficial role of abdominal belts and lum­ bar supports was inspired by the early theories of intra­ abdominal pressure (lAP) reducing spine compression forces. Because wearing a belt helps in generating higher IAP (Harman et al 1989, McGill et al 1990), it was assumed

Clinical biomechanics of the lumbar spine


that the belt was helpful in protecting the lumbar spine from excessive forces ( Harman et al1989, Lander et al1990, 1992). However, the literature to date does not support this notion. A very thorough and systematic literature review on lumbar supports by van Poppel et al ( 2000) demonstrated that many contTadictory results and findings of 'no effect' rule out most of the benefits with which the belts are often credited. For example, some studies found that abdominal belts mar­ force, spinal shrinkage and muscle EMG ( Bourne & Reilly ginally increase trunk strength; decrease spine compression

I f w e consider that only 1-2% of the maximum voluntary contraction ( MVC) is required from trunk muscles to main­ tain the spine in a stable upright posture ( Cholewicki et al 1997) ( see p. 72), the estimated belt effects might indeed be very small. An abdominal belt can enhance spine stability around its neutral posture by 40% at the most ( Cholewicki et al 1999a, Ivancic et aI 2002). Even if we assume full adap­ tation to this additional stability, the expected reduction in trunk muscle co-contraction will not exceed 0.8% MVC ( 40% x 2% MVC). Clearly, such small differences in muscle activation are beyond the detection accuracy of our current EMG recording techniques. Furthermore, based on a simple but realistic model of trunk flexors and extensors ( Cholewicki et al 1997), we can estimate the difference in spine compression force that corresponds to the 0.8% MVC reduction in muscle co-contraction to be roughly 35 N. Again, such a small reduction in the spine load appears nei­ ther statistically significant nor clinically relevant. Where then does the subjective perception of the benefits of wear­ ing an abdominal belt or a lumbar support come from? Let us examine a similar analogy to the one above. The addition of a 32 kg mass to the trunk requires an increase in trunk muscle co-contraction of approximately 1-2% MVC above the 1-2% MVC already required to maintain a stable upright posture of the spine without additional load ( Cholewicki et al 1997). Could, then, a reduction of 0.8% MVC be perceived as a relief equivalent to the removal of 12.8-25.6 kg from the upper body? Furthermore, sustained muscular contractions of 5% MVC or greater will eventu­ ally result in pain while the less intense contraction can be sustained indefinitely Gonsson 1978). Could it be that patients with low back pain, who exhibit more muscle co­ contraction during the activities of daily living ( Lariviere et al 2000, Marras et al 2001, van Dieen et al 2003), benefit from the reduction of muscle co-contraction below the 5% MVC threshold with the help of lumbosacral orthoses? Suggestions of improved trunk proprioception with lum­ bosacral orthoses have also been made ( McNair & Heine

1991, Granata et al 1997, Lee & Kang 2002, Smith et al 1996, Sullivan & Mayhew 1995, Warren et al 200 Woldstad & 1, Snook 1995, Ivancic et al 2002, Lantz & Schultz 1986a, Sherman 1998). Others found no such effects ( Ciriello &

Majkowski et al 1998, Marras et al 2000, McGill & Norman

1987, Rabinowitz et al 1998, Reyna et aI 1995). The positive findings are often related to the altered kinematics of task execution imposed by the belt and lower trunk moments, sion forces ( Granata et al 1997, Woldstad & Sherman 1998). which in tum result in misleadingly smaller spine compres­ The only consistent finding across various studies is that

belts reduce trunk range of motion and increase trunk stiff­ ness ( Axelsson et al 1992, Buchalter et al1988, Cholewicki et al 1999b, Fidler & Plasmans 1983, Lantz & Schultz 1986b, McGill et al 1994, Tuong et al 1998). Again, it has been shown mathematically that this increase in trunk stiffness translates directly into enhanced stability of the lumbar spine, even around its neutral posture ( Ivancic et al 2002). Thus, an abdominal belt and lAP can each individually or additively increase spine stability. Specifically, the estimates of these effects are as high as a 40% increase in spine stabil­ ity due to wearing a belt and another 40% due to generating large lAP for a combined effect from both mechanisms of more than an 80% increase in spine stability ( Cholewicki et al 1999b, Ivancic et al 2002). However, the difference between the two mechanisms is that the increase in spine stability due to high lAP is actively gained from muscle co-contraction associated withlAP. In contrast, the stabilizing effect of the belt is a passive mechanism stemming from the interaction of the wide and stiff belt placed between the ribcage and pelvis. Even though the spine stabilizing function of lumbar sup­ ports is relatively well documented, no objective clinically relevant benefits have been found. A prescription of abdom­ inal belts to manual load-handling workers does not reduce the incidence of low back injuries Gellema et al 200l, Reddell et al 1992, Wassell et al 2000). The efficacy of lumbosacral orthoses in the treatment of spine fractures or following a fusion surgery has not been completely proven ( Axelsson et al1995, Ohana et aI 2000). Even the concern of muscle weak­ unsupported in the literature ( Holmstrom & Moritz 1992, ening following long-term belt wearing appears to be

1999, Newcomer et al 2001). Perhaps enhanced propriocep­ tion in the lumbar spine may reduce the likelihood of low back injury and pain due to motor control error ( see section on stability based concept of musculoskeletal injury). Therefore, the perceived muscle weakening following long-term belt wearing might instead be motor control deconditioning. There are currently no data to help answer the above questions. For now, the identification of exact mechanisms underlying the sensation of the protective function of lumbar supports and back pain relief from wearing lumbosacral orthoses must await results from more theoretical and experimental studies.

Walsh & Schwartz 1990). However, many studies report that people perceive a sense of security and/or pain relief from wearing lumbar supports ( Ahlgren

Low back pain ( LBP) is a multifactorial problem. Numerous risk factors associated with acute low back injury and/or chronic disability have been identified. These risk factors

Alaranta & Hurri 1988, Million et al 198 1). P erhaps such mechanical effects are too small to be detected objectively.

& Hansen 1978,



fall into one of three major categories: demographic, such as an individual's strength and age; psychosocial, such as psy­ chological stress and job satisfaction; and biomechanical, such as posture and the load handled (Frank et al 1996). While their reported relative importance depends on the quality of the measurement tools used, it appears that the worst-case scenario is a combination of factors from all three categories (Kerr et al 2001). A well-designed preven­ tion or rehabilitation programme must take into account all three factors. The following sections will focus only on the biomechanical aspects of LBP and injury.






& Panjabi 1992).

Accumulation of end-plate fractures, which are often missed on conventional roentgenograms, and internal disc disruption have been proposed as the mechanisms leading to LBP and intervertebral disc degeneration (Schwarzer et al 1 995, van Dieen et aI 1999). Consistent with this model, exposure to both high cumulative and large peak spinal loads can lead to LBP (Norman et aI 1998). In addition to joint compression, the load borne by the spine includes shear forces and bending moments. Facets and the intervertebral disc support anterior joint shear force, which results mostly from the upper body weight and lifted load. Unless a spondylolysis or spondylolysthe­ sis is present, anterior shear force does not appear to be a threat to the integrity of the lumbar spine (Cyron et aI 1979). In vivo estimates of anterior shear of approximately 200 N (Potvin 1991) are well below the ultimate strength of the motion segments reported to be between 620 and 980 N (140-220 lb) (Miller et al 1986, Osvalder et aI 1993). P osterior ligamentous structures fail under relatively low loads when the lumbar spine is subjected to flexion moments (Adams et al 1980, Osvalder et al 1990). Thus, ruptured supraspinous and interspinous ligaments are commonly seen in adult spines (Grenier et al 1989, Rissanen 1 960). P osterior disc herniation is also associated with flex­ ion moments applied in the presence of a large spine com­ pression force (Adams & Hutton 1982a, 1982b). In addition, the ligaments and disc exhibit viscoelastic behaviour and creep when loaded during prolonged spine flexion. Peak lumbar flexion increased by 5.5 degrees after sitting for 20 minutes with fully flexed posture (McGill 1992b). Full recovery of spine mechanical properties took 30 minutes (McGill 1 992b). However, the neurophysiological response pathways between lumbar ligaments and muscies may not fully recover even after 7 hours (Jackson et al 2001, Solomonow et al 2002). Therefore, repetitive tasks per­ formed in flexed postures constitute a significant risk factor for overloading posterior ligaments and may lead to LBP. Despite the many identified biomechanical risk factors, the conventional model of musculoskeletal injury possesses several limitations that make it inconsistent with some doc­ umented circumstances of low back injury and LBP. Reported low back injuries in an occupational setting rarely involve near-maximum exertions (McGill 1997): An injury sustained during sub-maximal tasks is difficult to explain with the overload model when the same individual or oth­ ers performed the same task repeatedly in the past without any adverse effects. Sudden spine loading, trips and slips Frymoyer et al 1983, Manning et al 1984, Omino & Hayashi 1 992, Troup et a1 1981), but these scenarios may not neces­ sarily produce tissue loads that are above their physiologi­ cal limits. Finally, there is no consensus in the literature on the most detrimental biomechanical factors associated with LBP. Some researchers have identified peak loads while others have identified cumulative spine compression forces are also identified as causes of LBP (Bigos et al 1986,

Equi librium based concept of musculoskeletal injury
The conventional model of musculoskeletal injury is based on the concept of tissue overload during physical exertion. Tissue load tolerance is compared to the estimated loads in vivo. The injury is likely to occur if the tissue loads approach or exceed the tissue tolerance levels at any given time. This model encompasses several aspects of tissue fail­ ure such as the accumulation of microtrauma during repet­ itive exertions, tissue creep, fatigue and/or unbalanced tissue loading (Kumar 2001). The in vivo estimates of the compressive loads sustained by the lumbar spine during moderate physical exertions range between 2000 and 6000 N (450-1350 lb) (Davis et al 1 998, Potvin 1997, van Dieen et aI 2001). In the extreme case of competitive weightlifting, spine compression can reach 18 500 N (4150 lb) (Cholewicki et al 1991). On the other hand, the highest reported compressive load that a spinal motion segment withstood to failure during in vitro tests was just under 16 000 N (2900 lb) (Hutton et al 1979). On average, specimens fail under loads of approximately 6000 N (1350 lb) (Brinckmann et al 1989, Granhed et al 1 989, P orter et aI 1989). This apparent paradox of incompatibility between in vivo spine loads and in vitro tolerance levels motivated the formulation of several spine-unloading theo­ ries. The mechanisms involving intra-abdominal pressure (discussed in the section on the role of intra-abdominal pressure in spine stability), lumbodorsal fascia as the hydraulic amplifier, and the posterior ligamentous system have been proposed (Gracovetsky et al 1 985, 1989, 1990). These hypotheses found very little support from the stud­ McGill 1 992, McGill & Norman 1988), but unfortunately many of the recommendations derived from these theories are still being perpetuated. Direct comparisons between in vivo and in vitro failure loads are ill-advised because the cadaveric specimens are generally harvested from individuals older than the popu­ lations used in in vivo studies (Brinckmann et al 1 989, Granhed et al 1989, P orter et al 1 989). The specimens are frequently degenerated and have less bone mineral content, related often to prolonged bed rest or illness. On the other hand, sub-failure injuries can occur much earlier, before the ies that followed (Adams & Hutton 1 986, Cholewicki &

Clinical biomechanics of t h e lumbar spine


as the pertinent risk factors ( Kerr et al 200 Norman et al 1, 1998, van Dieen et a1 2001). Shear forces, excessive bending and twisting, the frequency of movement and whole body vibration have also been proposed as risk factors ( Damkot et al 1984, Kelsey et al 1984, Kerr et al 2001, Manning et al 1984, Marras et al 1995, Pope et al1998). It appears that any activity requiring physical exertion constitutes a risk factor for sustaining low back injury. Therefore, not all of these data are consistent with the model of tissue overload pre­ sented above. However, an injury model based on spine instability may better explain the above findings.

The motor control of spine stability i s extremely com­ plex. If we assume 5 degrees of freedom at each interverte­ bral joint ( three axes of rotation and anteroposterior and lateral translations), the entire lumbar spine will comprise 30 degrees of freedom ( 5 x 6 joints). With a multitude of muscles and redundant lines of action, there exists an infi­ nite number of possible muscle activation patterns that will satisfy equilibrium constraints, but an adequate stability level may not necessarily be achieved. P roblems of motor control and stability of the lumbar spine constitute an extension of the traditional equilibrium based approach to musculoskeletal injury. To date, very few spine stability studies have been published and they are McGill 1996, Gardner-Morse et al 1995, Granata & Marras limited to static conditions ( Bergmark 1989, Cholewicki &

Stability based concept of musculoskeletal injury
A stability based model of spine injury was first proposed by Panjabi ( 1992a). He identified three subsystems: the pas­ sive subsystem consisting of ligamentous structures and disc; the active subsystem consisting of muscles; and the motor control coordinating the fulfilment of stability demands between the other two subsystems. A variety of mechanoreceptors, including but not limited to muscle spindles, Golgi tendon organs, joint receptors and cuta­ neous receptors, provide continuous feedback to the motor control system. A dysfunction in any of these subsystems may result in or lead to a clinical problem and/or it must be Cholewicki & McGill ( 1996) extended this model further and quantified the stability of the lumbar spine given its posture, external loads and trunk muscle activation ( EMG). They demonstrated that spine instability or buckling could occur if the level of muscle co-contraction is low or their activation pattern is erroneous. Furthermore, Cholewicki & McGill ( 1992) observed a minor injury via fluoroscopy of a power lifter executing an extremely heavy lift. A hyperflex­ ion at only one intervertebral level ( L4-5) occurred during the lift suggesting a buckling phenomenon of the lumbar spine. Thus, the above studies highlighted motor control error as a possible factor precipitating low back injury and pain. compensated by the remaining subsystems ( Fig. 7.8).

2000). Nevertheless, these recent efforts have opened new horizons for understanding spine disability and LBP. Based on stability analyses, it is now possible to explain several phenomena that traditional approaches have been unable to adequately elucidate. New hypotheses regarding spine injury mechanisms were formulated and tested. The fol­ lowing sections explore certain features of this model in more detail and in this context review the research related to muscle recruitment pattern and motor control in healthy individuals and in patients with mechanical LBP.

Explanation for in jury occurrence under very low loads
Situations when individuals 'throw out their back' when picking up small objects from the floor or tying their shoelaces are common. Traditional equilibrium modelling does not provide an adequate explanation for such phe­ nomena. Stability, on the other hand, offers much insight into possible injury mechanisms. Light tasks requiring little muscular effort create a scenario in which the spine is most vulnerable to buckling ( Cholewicki & McGill 1996). In these situations, muscular fatigue or a motor control error may lead to spine instability. To prevent spine buckling, small intervertebral muscles that bridge an unstable lumbar level must be activated. Independent recruitment of large mus­ cles that span several lumbar levels may not be a suitable response, as these muscles increase the compressive load on the spinal column. Their activation would increase the buckling effect, if unaccompanied by activation of small intervertebral muscles. Consequently, small muscles and passive supporting structures may be overloaded and injured or joint instability may result in abnormal motions which would irritate soft tissues, nerve roots or nociceptors. As discussed on p. 72, co-activation of 1-2% MVC of trunk flexors and extensors is present and necessary to assure the mechanical stability of the spine in an upright posture ( Cholewicki et al 1997). This level of muscle co­ activation must be maintained throughout the duration of an entire day when individuals are walking or sitting.

Figure 7.8 Panjabi's mode l of spina l stabi lity and its motor con­ tro l. Adapted with permission fro m Panjabi (1992).

A two-fold increase in trunk muscle co-contraction was necessary to maintain spine stability when stiffness of



contribution of the passive subsystem was reduced in a biomechanical model (Cholewicki et al 1997). This decrease in passive subsystem stiffness can be the result of mechanical trauma or a sub-failure injury (Oxland & Panjabi 1992). Because sustained muscular contractions at the level of 5% MVC or greater lead to muscular fatigue and pain (Jonsson 1978), the co-activation of trunk mus­ cles during upright standing should be well below the 5% MVC value. Consequently, if decreased passive stiffness or motor control dysfunction exists, these muscles may increase activation and become fatigued, resulting in an inability to provide the adequate degree of spine stability when attempting certain physical tasks. These events may lead to a vicious cycle in which the spine becomes repeat­ edly re-injured because of muscle fatigue. Clinically, increased levels of muscle co-activation may indicate dys­ function of the passive stabilizing system of the lumbar spine. A similar hypothesis was first proposed by Panjabi ( 1992a, 1992b). This serves as a plausible explanation for chronic mechanical LBP. There is also evidence of poor position sense, diminished postural control and slow reaction times in patients with mechanical LBP (Oddsson et al 1999, Taimela et al 1999, Wilder et al 1996). Certainly, if trunk stability is compro­ mised by abnormal patterns of muscle activation or poor postural control it leaves the spine vulnerable to injury, especially under sudden loading conditions. A motor con­ trol problem fits with an instability /motor control model of low back injury, which overcomes many limitations of the conventional model. Using the instability/motor control model, injuries that occur at low effort levels such as a bending movement, twisting or reaching for an object can finally be explained.

1994, Edgerton et al 1996, Hodges & Richardson 1996,

dysfunction can be identified (Cresswell & Thorstensson

Mannion & Dolan 1994, O'Sullivan et a1 1997a, Paquet et al clinical diagnosis of lumbar instability appear to preferen­ tially activate the rectus abdominis and/or external oblique muscle groups (O'Sullivan et al 1997a, 1998, Silfies 2002). These patterns of muscle activation were interpreted as a dysfunction of the transversus abdominis and lumbar mul­ tifidus muscle groups in providing adequate compensation for a mechanically compromised osteo-ligamentous spine or passive subsystem. However, others did not find such a pattern in LBP patients (van Dieen et al 2003). Two models have been proposed in the past to explain different muscle recruitment patterns in patients with LBP. The pain-spasm-pain model postulates that pain results in increased muscle activity, which in turn will cause pain (Roland 1986). The pain-adaptation model states that pain decreases the activation of muscles when active as agonists and increases it when the muscle is active as antagonist (Lund et al 1991). The effects of such a control strategy would be that movement velocity is reduced and move­ ment excursions are limited. Both theories yield conflicting predictions on how LBP patients would alter trunk muscle recruitment in response to their pain, yet both find some supportive evidence in the literature. Recent work by van Dieen et al (2003) demonstrated a higher lumbar to thoracic erector spinae activation ratio and a greater level of trunk muscle co-contraction in a LBP group compared to asymptomatic controls. These EMG data were then fed into a biomechanical model (Cholewicki

1994, Peach et a1 1998, Sihvonen et aI 1991). Patients with a

& McGill 1996), which indicated that this change in recruit­
ment pattern enhanced spinal stability (van Dieen et al 2003). These authors suggested that patients might be uti­

Muscle recruitment patterns and low back pain
Biomechanical modelling of lumbar spine stability clearly identifies antagonist muscle co-activation as a mechanism by which the entire spinal column becomes stiffer, hence more stable (see p. 72). It has been suggested that 25% MVC of the trunk musculature provides maximal trunk stiffness (Cresswell & Thorstensson 1994). Even larger levels of

lizing many different muscle recruitment patterns with a cornmon goal of enhancing spinal stability.

Motor control of spine stability and low back pain
Due to the multisegrnental structure of the human body, any voluntary movement is associated with postural adjustments. Thus, control of balance and lumbar stability are essential requirements for pain-free function of the spine. Motor control operates through the integration of several different pathways. Spinal pathways use proprio­ ceptive input from sensory organs, muscles and joint struc­ tures to assist in postural control and trunk stability. The peripheral sensory system (spinal reflex pathways) also functions in conjunction with brain stem and cognitive pro­ gramming. The brain stem coordinates visual, vestibular and joint receptor information, while cognitive program­ ming is based upon repeated or stored central commands. The functional assessment of trunk motor control related to the maintenance of spinal stability is difficult owing to the complexity of this system and the continually changing demands for stability and movement. Motor control research related to spine stability has been accomplished

trunk muscle co-activation may be necessary to stabilize the lumbar spine during more complex and dynamic tasks muscle co-activation functions to increase spinal stability Cresswell et al 1994, Gardner-Morse & Stokes 1998, Gracovetsky et al 1985) and by providing compressive loads to the spinal column (Janevic et al 1991, Stokes et al 2002). It is not surprising, then, that a number of studies have reported more antagonistic muscle co-contraction during various activities in patients with LBP (Lariviere et a1 2000, Marras et a1 200l, van Dieen et aI 2003). In general, inconsistent differences in trunk muscle recruitment patterns in patients with mechanical LBP have been reported and thus, no particular pattern of muscle by increasing muscle stiffness (Cholewicki et al 1999a, (Lavender et al 1992a, Marras & Mirka 1996). Antagonist

Clinical biomechanics of the lumbar spine


predominantly through monitoring of EMG activation pat­ terns (synergist and antagonist), postural control parame­ ters and muscle onset and offset timing. Several models of testing muscle response to a controlled challenge have been established: 1. use of anticipated self-perturbation of the extremities (Hodges & Richardson 1996, 1997b, Zattara & Bouisset 1988) 2. use of expected or unexpected external loading or loading of the trunk (Radebold et al 2000, van Dieen & de Looze 1999a, Wilder et a1 1996) 3. standing or seated balance control (Mien*s & Frank 1999, Radebold et al 2001, Takala et a1 1997) 4. use of forced or altered breathing patterns (Hamaoui et al 2002, McGill et a1 1995) 5. use of expected or unexpected perturbation of a support surface (Huang et aI 2001). Postural adjustments triggered prior to the onset of vol­ untary movements appear variable and task specific in asymptomatic individuals (Andersson et al 1995, Oddsson et al 1999). It has been demonstrated that combinations of planned tasks with unexpected perturbation could cause some conflict between the two commands that may increase the risk of injury or motor control errors (Oddsson et al 1999). In addition, pain or prior injury to muscu­ loskeletal tissues containing mechanoreceptors may also provide inaccurate information to the motor control system creating a mechanism for motor control errors and further injury to musculoskeletal tissue (De Luca 1993, Hodges & Richardson 1998, Mienljes & Frank 1999, Radebold et al 2000, Solomonow et al 2001, Takala et aI 1997). Through analysis of asymptomatic individuals during self-perturbation of an extremity, the transversus abdo­ minis (TrA) and internal oblique (IO) have been identified as acting in a feed-forward or preparatory manner (Hodges & Richardson 1996, 1997b, 1999, Hodges et aI 1999). It also appears that activation of the TrA and 10 may be a general response to disturbance of the centre of mass, as their acti­ vation was not direction or movement specific (Aruin & Latash 1995, Hodges & Richardson 1997a). This prepara­ tory activation of the TrA may contribute to control of spinal segmental motion, which theoretically is necessary to prepare the spine for contraction of other musculature. It follows from this discussion that the trunk musculature would require appropriate recruitment and timing to main­ tain stability of the spine during static posturing and movement (Cholewicki et al 1997, Gardner-Morse & Stokes 1998, Hodges & Richardson 1996). In turn, this would require accurate and timely information from the mechanoreceptors in the spine to allow for appropriate adjustments of the trunk musculature via the motor control system to maintain spinal stability. A number of studies compared postural control of asymp­ tomatic individuals to patients with LBP. Results of studies employing unilateral self-perturbation of the limbs suggest

that there is a dysfunction in the motor control system related to delayed activation of the transverse abdominis muscle group in chronic LBP subjects. This delayed activa­ tion of the TrA could be a contributing factor to the inability to stabilize the spine (Hodges 2001, Hodges & Richardson 1996, 1997b). In a sudden trunk loading paradigm, patients with LBP demonstrated delayed onset latencies of trunk muscles. In addition, LBP subjects responded with a pattern of trunk muscle co-contraction instead of the selected direc­ tional response utilized by healthy subjects (Magnusson et al 1996, Radebold et al 2000, 2001, Wilder et aI 1996). These pro­ longed latencies and co-contraction patterns may represent a motor control adaptation for enhancing lumbar stability or an impairment making it difficult for patients to cope safely with sudden and unexpected loading. Impairments in standing postural control have been reported in patients with LBP (Mien*s & Frank 1999, Takala et al 1997). Increased body sway has been related to dys­ function in proprioception stemming from damage or injury to lumbar spine tissue containing mechanoreceptors. Similar findings were reported for sitting balance, with LBP patients performing significantly poorer especially with increased seat instability and lack of visual feedback (Radebold et al 2001). This finding appears to support the notion that pro­ prioceptive input is somehow altered in patients with LBP, as absence of visual feedback increases the challenge to pos­ tural control. Significant correlations between poor sitting balance with eyes closed and longer trunk muscle response latencies to a sudden load release (Radebold et a1 2001) sup­ port the hypothesis that altered gross motor control stems from nociceptive stimuli or poor proprioception. This hypothesis is further supported by studies that have docu­ mented poor lumbar position sense (Gill & Callaghan 1998, Parkhurst & Burnett 1994, Taimela et a1 1999) and longer psy­ chomotor reaction speed (Luoto et al 1996, 1999, Taimela et al 1993) in patients with mechanical LBP. Thus, studies testing spinal reflexes and brain stem pathways of the motor control system reveal alterations of both the feed-forward and feed­ back neuromotor strategies in patients with LBP.
Cause or effect?

While it is well documented that differences in motor con­ trol parameters do exist in individuals with mechanical LBP, it is not known at this time whether these differences are the cause or effect of LBP. Longitudinal prospective studies are necessary to answer this question, but to date none have been published. Impaired proprioception in the lumbar spine, delayed trunk muscle reflex response and poor postural control may represent predisposing factors to the development of LBP by hindering proper responses to dynamic loading and fail­ ure to provide adequate stability to the spine. Individuals susceptible to LBP could inherently possess those risk fac­ tors or acquire them after the first episode of back injury (Fig. 7.9). For example, the subjects used in a majority of the studies were classified as having chronic LBP and may



within this subsystem. Therefore, the next question arises,



Effect? Changes in motor control

as to which motor control alterations constitute beneficial adaptation and which are a detrimental impairment.

Impairment or adaptation?
The differences in muscle recruitment or neuromuscular control seen between patients with LBP and asymptomatic individuals have been hypothesized to be either (a) a com­
'Non-copers'? 'Capers'?

pensation for underlying spinal instability, passive struc­ ture damage or proprioceptive dysfunction (Lariviere et al 2000, Radebold et a1 2000, van Dieen et a1 2003), or (b) some impairment predisposing these patients to sustain recurrent

Figure 7.9 A diag ra m of the re lationship between low back pain ( L BP) o r inju ry ( LBI) and moto r cont ro l changes docu mented in lit­ e rature. It is cu rrently not known whethe r the dif fe rences in moto r cont ro l in LBP patients a re the cause o r effect of LBP. Fu rthe rmo re, one of the most i mpo rtant c linica l questions is which changes con­ stitute functiona l adaptation and which are i mpai rment ('cope rs' v s 'non-cope rs')

injuries (Cholewicki et a1 2002b, Hodges & Richardson 1996, O'Sullivan et al 1997a, 1998, Radebold et al 2000, Sihvonen

et al 1997). Correct classification of patients based on the above possibilities is a critical step for the selection of effec­ tive therapy (Fig. 7.9). Perhaps interpretation of the changes in motor control depends on many individual factors in a particular patient. For example, in someone with acute LBP this altered pattern may display hyperactivity or inhibition

represent those who are unable to develop appropriate mechanisms via the active and motor control subsystems to allow for pain-free function of the spine ('non-copers'). On the other hand, one study that reported similar motor con­ trol changes in first-time injured athletes in spite of their clinical and functional recovery raised the possibility of a chronic condition being a series of acute events (Cholewicki et al 2002b). The recovery of damaged mechanoreceptors and in turn motor control may take longer than functional recovery and subsidence of pain. This impairment, in turn, can further predispose an individual to sustain recurrent low back injuries. In fact, previous back trouble appears to be the best predictor of future LBP (Bigos et al 1991, Feyer et a1 2000, Greene et a1 200l, Schneider et a1 2000). The changes in motor control observed in LBP patients could also result from LBP. They could function as a com­ pensation mechanism designed to stabilize the lumbar spine following injury or may be an impairment caused by LBP (Fig. 7.9). Damage or inflammation in tissues contain­ ing mechanoreceptors could alter their feedback and in turn impair motor control. Finally, changes in muscle recruitment pattern could also result from inhibition or hyperactivity of specific muscles due to pain (Edgerton et al 1996) or be caused by pain itself (Arendt-Nielsen et al 1996, Cobb et al 1975, Sterling et a1 200l, Zedka et al 1999). To our knowledge, only one prospective study addressed the causality issue of ankle joint instability and postural control (Tropp et al 1984). These authors found that poor performance in a postural control task resulted in a significantly higher risk of sustaining ankle sprain injury among professional soccer players. Thus, these results would suggest that impaired motor control is the cause of injuries, although extrapolation from the ankle joint to the spine is uncertain. In either case, intervention based upon restoring a functional and adequate motor control strategy may be beneficial to individuals demonstrating alterations

secondary to pain, while in a chronic (non-inflammatory) LBP, such an alteration may suggest an inadequate adaptive response in an attempt to enhance stability over a multitude of tasks ('non-coper '). Answering this question would require follow-up studies of acute LBP patients and monitor­ ing muscle activation patterns of chronic LBP patients across a large number of tasks and conditions. What changes constitute an impairment and which are an adaptation? From a biomechanical perspective, this piece of the puzzle is currently missing and is difficult to define with­ out a 'standard motor control pattern' or an in vivo measure­ stability or stiffness of the trunk in relation to the trunk motor ment of spine stability. Methods that allow quantification of control pattern are in the early stages of development. Once

this quantification is achieved, the interpretation is still diffi­
cult. If trunk stiffness or stability in LBP patients is higher than in asymptomatic individuals, does that mean they are

co-contracting too much and creating excessive spine com­ decrease muscle activity or achieve skilled co-contraction

pression forces? If this is the case, perhaps assisting them to

strategies that can provide increased trunk stability without the excessive compression penalty is the right course of action. However, if LBP patients actually require that much significant tissue damage and adequate adaptation ('copers'). the best intervention. From a clinical perspective, the answer may be as fol­ lows: if an alteration in muscle activation or motor control allows the individual to function in daily activities, at work or at play, we might label this alteration as an adap­ tation. However, if this individual is demonstrating func­ tional limitations and/or disability, we may label this pattern as an impairment. In clinical practice, we are inclined to lean towards the impairment label, as most patients are seeking assistance because of pain, functional co-contraction to maintain spine stability, this may indicate

In this case, altering the co-contraction strategy may not be

Clinical biomechanics of the lumbar spine


limitation and/ or disability. We therefore assume that they are protecting injured structures, avoiding nociceptive stimulation or are unable to adequately compensate for their dysfunction using their present motor control strate­ gies and are thus impaired. Clinically, acuity of symptoms along with other clinical measurements of physical impair­ ments and function guide our decision related to interven­ tion with a particular patient. Thus, lumping all altered muscle activation patterns or motor control changes into either an adaptation or an impairment may be a gross mis­ interpretation of both clinical and research findings.

Review of the current research related to trunk motor control reveals considerable variability in co-contraction strategies, activation patterns and timing of muscle activation in the asymptomatic population. In part, this has created some dif­ ficulties in the research arena related to determining 'stan­ dard' motor control strategies. In some ways, this variability should be expected because of the redundancy of the trunk musculature and complexity of the motor control system (Latash et al 2002). If we take skilled golfers for example, and compare their swings, we would find they generally adhered to a pattern of motion, but with slight variations in joint range, trajectory, segment coordination and timing. Yet these individuals still accomplish the same task with relatively equal skill. In much the same way, more than one co­ contraction, activation or timing pattern may be capable of achieving adequate spinal stabilization. Similar findings are reported in the literature related to knee instability, where no single 'good compensation' strategy was adopted by patients with anterior cruciate ligament injury (Rudolph et al 1998).
Cli n i cal assessment of trunk stability and motor control

Ideally, during the evaluation of a patient with LBP, the clinician attempts to determine the presence or absence of potential factors that may be contributing to mechanical low back dysfunction. These factors are then used to establish a diagnosis and treatment plan. The present limitation to this clinical decision making process as it relates to the spinal instability / motor control model of LBP is that clinical tests for the evaluation of trunk motor control (muscle recruit­ ment patterns, proprioception and postural control) are in their infancy. Evaluation of muscle activation patterns has recently achieved some attention based primarily on the work of O'Sullivan, Richardson, Jull and colleagues Gull & Richardson 2000, Jull et al 1993, O'Sullivan 2000, O'Sullivan et al 1997a, Richardson & Ju1l 2000). Assessment of trunk sta­ bility during self-perturbation of the extremities has been proposed by Van Dillen and co-workers (Van Dillen et al 2001, 1998). This assessment technique uses observation of spine kinematics, muscle palpation and symptom repro­ duction in several different trunk positions (sitting, lying and standing). If patients are unable to maintain a neutral lumbar position while performing self-perturbation, the clinician hypothesizes that a motor control deficit exists. A review of the literature would also suggest that assess­ ment of trunk proprioception or sitting balance, particularly without visual feedback, might provide evidence of motor control dysfunction (Radebold et a1 2001). To our knowledge, these types of clinical assessment techniques are at the forefront of current LBP research and have yet to be systematically developed and tested for validity and reliability in diagnosing motor control dys­ function. To date, our ability in most cases to make a clear clinical diagnosis of a motor control dysfunction in LBP patients is limited to many assumptions. For further dis­ cussion of clinical examination techniques, we refer you to the current research and chapters 10, 22 and 31 in this text on lumbar spine motor control.
I mplicati ons for rehab i l itation strategies

One clinical problem is identifying those patients with mechanical low back pain who would most benefit from a motor control training approach, as LBP results from a com­ bination of factors. Our current inability to determine which impairments are contributing to an individual's mechanical LBP has been an obstacle within the clinical community. Since most LBP patients present with multiple impairments, we have acquiesced in treating them with a multifaceted approach. The routine rehabilitation pro­ gramme for a patient diagnosed with mechanical LBP may consist of bracing, lower extremity muscle stretching, trunk muscle strengthening and endurance exercises, postural exercises, dynamic stabilization exercises, general condi­ tioning exercises, modalities for reduction of pain and inflammation and education in proper lifting techniques. At present, treatment is essentially global because it is unclear which particular interventions help improve indi­ vidual patient outcomes.

Despite our inability to determine whether motor control differences are a risk factor for the development of LBP or the effect of injury and pain, a treatment approach for mechanical LBP has been developed based on Panjabi's model (Panjabi 1992a). According to this model, the muscu­ lar and motor control subsystems are trained to 'appropri­ ately' control and stabilize the spine (Fritz et al 1998, Norris 1995, O'Sullivan et al 1997b, Richardson & Jull 2000, Saal & Saal 1989). Several studies have demonstrated the benefits of addressing motor control in the treatment of LBP. Patients receiving treatment programmes directed toward enhancing motor control demonstrated significantly less pain, a faster return to function, and had fewer reoccur­ rences of LBP at follow-up (Hides et a1 2001, O'Sullivan et al 1997b, 1998, Sihvonen et al 1997). Thus, it may be possible to train the motor control system to provide sufficient dynamic stability to a mechanically compromised lumbar



spine (Hides et al 200 1). What remains inconclusive is whether such treatment truly improves the parameters of motor control such as muscle reaction times or patterns of activation. Improvements with these protocols may be due to other effects of training such as increased muscle strength or endurance, mood elevation, biochemical changes or modulation of pain. Only one study to date has demon­ strated improved reaction times to match those of healthy control subjects during unexpected perturbation following a specialized rehabilitation programme (Wilder et al 1996). According to the spine stability/motor control model and given the fact that all trunk muscles contribute to appear that training of the entire neuromotor apparatus might be more beneficial than focusing on individual mus­ cle training. Given the variability of the motor control sys­ tem (Latash et al 2002) and the redundancy in the trunk musculature, there may be more than one effective muscle activation pattern with which spine stability can be achieved. Recently, several rehabilitation strategies based 1995, O'Sullivan 2000, Richardson & Jull 2000, Saal & Saal 1989, Saal et al 1990). The aim of these strategies is to help individuals to develop better control of the trunk muscles so that they can be adequately recruited during physical activities. The lack of a 'gold standard' compensatory mus­ cle activation strategy creates complications for designing treatment programmes to improve lumbar spine stability. As such, successful training strategies have to provide the opportunity for development of individualized compensa­ tory patterns of the trunk musculature. This raises some questions regarding the effectiveness of programmes that emphasize one specific motor control training pattern. Another aspect of a rehabilitation programme is the intensity of exercise. The research suggests that trunk mus­ cle co-contractions at 1-2% MVC for a healthy spine, 2-5% MVC for a compromised spine or at most 10-25% MVC Cresswell & Thorstensson 1994) are sufficient to stabilize the spine. Thus, traditional strengthening protocols (high load, low repetitions) may not be necessary to achieve ade­ quate spine stabilization over the course of daily activities. Because large muscle forces are not typically required for daily function, it would appear that effective spine stabi­ lization requires the ability to co-contract trunk muscles at low levels over long periods of time and under a variety of postures and tasks. In addition, we argued earlier that cir­ cumstances involving sudden spine motion and lower loads leave more room for motor control errors. Thus, the exercise prescription should lean toward the parameters of muscle endurance (low load, high repetitions), with emphasis on dynamic not static endurance activities. Muscle timing and postural control are also important factors to maintaining appropriate spine stability particu­ larly in the event of support surface unsteadiness and sud­ den or unexpected loading. Thus, dynamic exercises that (Cholewicki on 'stabilization training' have been introduced (Norris spinal stability (Cholewicki & VanVliet 2002), it would

challenge these particular parameters would be an impor­ tant component of motor control rehabilitation. Again, we believe that these exercises should be completed in a way that allows the patient to develop their own stabilization strategies. This follows the line of intervention being pro­ posed and tested regarding the rehabilitation of individuals Davies & Dickoff-Hoffman 1993, Eils & Rosenbaum 2001, et a1 2002). One would also expect that a learning process exists that may start with patients responding to these dynamic situa­ tions with gross co-contraction of the trunk musculature and eventually progressing to more skilled co-contraction patterns to achieve the desired control and stability. The motor learning theory of Bernstien hypothesizes that initial solutions to motor control problems result in 'freezing out' a portion of the degrees of freedom (Vereijken et al 1992a, 1992b). This 'freezing out' could be accomplished by keep­ ing the joints or segments rigidly fixed, allowing little to no motion, or by coupling of several degrees of freedom to form a joint complex. Improvement in skill would then be characterized by gradually reducing gross co-contraction or freeing degrees of freedom and moving towards compen­ satory synergistic muscle patterns during dynamic activi­ ties (Vereijken et al 1992a, 1992b). Evidence of a gross co-contraction strategy in mechanical LBP subjects has been reported by several investigators (Lariviere et al 2000, Marras et a1 2001, Radebold et a1 2000, van Dieen et a1 2003). Further discussion of stabilization exercises, motor con­ trol training programmes and recommended progression is contained in chapters 22 and 31 in this text. Concerns related to spine compressive and shear forces arising with muscle co-contraction exercises were addressed in the recent research by several authors (Allison et al 1998, Callaghan et al 1998, McGill 1998, Vera-Garcia et a1 2000). The motor control assessment and treatment techniques described in this chapter are in their relative infancy. Further controlled studies are required to determine their diagnostic and prognostic value and the treatment efficacy they afford. Only recently, research tools have been devel­ oped to test the model of low back injury and pain based on hypotheses spawned from this model have alread y charted new directions in the prevention, diagnosis and rehabilita­ tion of low back pain. motor control of lumbar spine stability. However, the Arokoski et al 1999, 2001, 2002, Axler & McGill 1997, with ankle, knee and shoulder instabilities (Beard et al1994,

Fitzgerald 1998, Maitland et al 1999, Rozzi et al 1999, Wilk

& McGill 1996, Cholewicki et al 1997,

lumbar spi ne biomecha nics stability motor control low back pa in

C l i n i ca l b i o me c h a n ics of t h e l u m ba r s p i n e


The authors would like to acknowledge their financial support from the National Institutes of Health, grant lR01 AR 46844-01 A l .

Adams M A, Hutton W C 1982a The mechanics of prolapsed Adams M A, Hutton W C 1982b Prolapsed intervertebral disc: a hyper­ Adams M A; Hutton W C 1986 Has the lumbar spine a margin of safety Adams M A, Hutton W C, Stott J R 1980 The resistance to flexion of the Ahlgren S A, Hansen T 1978 The use of lumbosacral corsets prescribed Alaranta H, Hurri H 1988 Compliance and subjective relief by corset treatment in chronic low back pain. Scandinavian Journal of Rehabilitation Medicine 20: 133-136 lumbar intervertebral joint. Spine 5: 245-253 in forward bending? Clinical Biomechanics 1: 3-6 flexion injury. 1981 Volvo Award in basic science. Spine 7: 184-191 intervertebral disc. International Orthopaedics 6: 249-253 Bergmark A 1989 Stability of the lumbar spine: a study in mechanical Bigos S J, Spengler D M, Martin N A et al 1986 Back injuries in engineering. Acta Orthopaedica Scandinavica 230(Suppl.): 1-54

Bigos S J, Battie M C, Spengler D M et al 1991 A prospective study of work perceptions and psychosocial factors affecting the report of back injury [published erratum appears in Spine 1991 16(6): 688].

industry: a retrospective study. II: Injury factors. Spine 1 1 : 246-251

Bourne N D, Reilly T 1991 Effect of a weightlifting belt on spinal Brinckrnann P, Biggemann M, Hilweg D 1989 Prediction of the 606-610 shrinkage. British Journal of Sports Medicine 25: 209-212

Spine 16: 1-6

for low back pain. Prosthetics and Orthotics International 2: 101-104

compressive strength of human lumbar vertebrae. Spine. 14:

Allison G T, Godfrey P, Robinson G 1998 EMG signal amplitude Electromyography and Kinesiology 8: 51-57

Buchalter D, Kahanovitz N, Viola K, Dorsky S, Nordin M 1988 Three­ dimensional spinal motion measurements. 2: A noninvasive Spinal Disorders 1 : 284-286 assessment of lumbar brace immobilization of the spine. Journal of

assessment during abdominal bracing and hollowing. Journal of

Andersson E, Oddsson L, Grundstrom H, Thorstensson A 1995 The role of the psoas and iliacus muscles for stability and movement of the Science in Sports 5: 10-16 lumbar spine, pelvis and hip. Scandinavian Journal of Medicine and

Callaghan J P, Gunning J L, McGill S M 1998 The relationship between Physical Therapy 78: 8--1 8 lumbar spine load and muscle activity during extensor exercises.

Arendt-Nielsen L, Graven-Nielsen T, Svarrer H, Svensson P 1996 The during gait: a clinical and experimental study. Pain 64: 231-240 influence of low back pain on muscle activity and coordination

Cherkin D C, Deyo R A, Volinn E, Loeser J D 1992 U se o f the

hospitalizations for mechanical low back problems in administrative Cholewicki J, McGill S M 1992 Lumbar posterior ligament involvement measurements. Journal of Biomechanics 25: 1 7-28 during extremely heavy lifts estimated from fluoroscopic databases. Spine 17: 817-825

International Classification of Diseases (ICD-9-CM) to identify

Arokoski J P, Kankaanpaa M, Yalta T et al 1999 Back and hip extensor Medicine and Rehabilitation 80: 842-850 muscle function during therapeutic exercises. Archives of Physical

Arokoski J P, Yalta T, Airaksinen 0, Kankaanpaa M 2001 Back and of Physical Medicine and Rehabilitation 82: 1089-1098

abdominal muscle function during stabilization exercises. Archives

Cholewicki J, McGill S M 1994 EMG assisted optimization: a hybrid biomechanical model. Journal of Biomechanics 27: 1287-1289 approach for estimating muscle forces in an indeterminate

Arokoski J P, Yalta T, Kankaanpaa M, Airaksinen 0 2002 Activation of paraspinal and abdominal muscles during manually assisted and non-assisted therapeutic exercise. American Journal of Physical Medicine and Rehabilitation 81: 326-335

Cholewicki J, McGill S M 1995 Relationship between muscle force and Journal of Biomechanical Engineering 117: 339-342 stiffness in the whole mammalian muscle: a simulation study.

Aruin A S, Latash M L 1995 Directional specificity of postural muscles movements. Experimental Brain Research 103: 323-332 in feed-forward postural reactions during fast voluntary arm

Cholewicki J, McGill S M 1996 Mechanical stability of the in vivo Clinical Biomechanics 1 1 : 1-15

lumbar spine: implications for injury and chronic low back pain.

Ashton-Miller J A, Schultz A B 1991 Spine instability and segmental use of terms. Seminars in Spine Surgery 3: 136-148

hypermobility biomechanics: a call for the definition and standard

Cholewicki J, VanVliet J Jr 2002 Relative contribution of trunk muscles to the stability of the lumbar spine during isometric exertions. Clinical Biomechanics 17: 99-105

Axelsson P, Johnsson R, Stromqvist B 1992 Effect of lumbar orthosis on Spine 17: 678-681 intervertebral mobility: a roentgen stereophotogrammetric analysis.

Cholewicki J, McGill S M, Norman R W 1991 Lumbar spine loads in Sports and Exercise 23: 1 1 79-1186 during the lifting of extremely heavy weights. Medicine and Science

Axelsson P, Johnsson R, Stromqvist B, Nilsson L T, Akesson M 1995

Cholewicki J, McGill S M, Norman R W 1995 Comparison of muscle forces and joint load from an optimization and EMG assisted lumbar spine model: towards development of a hybrid approach.

Axler C T, McGill S M 1997 Low back loads over a variety of Medicine and Science in Sports and Exercise 29: 804-811

8: 284-288

value in 50 cases followed prospectively. Journal of Spinal Disorders

Orthosis as prognostic instrument in lumbar fusion: no predictive

Cholewicki J, Panjabi M M, Khachatryan A 1997 Stabilizing function of 22: 2207-221 2

Journal of Biomechanics 28: 321-331

abdominal exercises: searching for the safest abdominal challenge.

trUllk flexor-extensor muscles around a neutral spine posture. Spine

Bartelink D L 1957 The role of abdominal pressure in relieving the Joint Surgery (British volume) 39: 718--725

pressure on the lumbar intervertebral discs. Journal of Bone and

Cholewicki J , Juluru K , McGill S M 1999a Intra-abdominal pressure Biomechanics 32: 13-17 mechanism for stabilizing the lumbar spine. Journal of

Bean J C, Chaffin D B, Schultz A B 1988 Biomechanical model programming method. Journal of Biomechanics 21: 59-66 calculation of muscle contraction forces: a double linear

Cholewicki J, Juluru K, Radebold A, Panjabi M M, McGill S M 1999b

Beard D J, Dodd C A, Trundle H R, Simpson A H 1994 Proprioception

enhancement for anterior cruciate ligament deficiency: a prospective Joint Surgery (British volume) 76: 654-659 randornised trial of two physiotherapy regimes. Journal of Bone and

Cholewicki J, Simons A P D, Radebold A 2000 Effects of external trUllk loads on lumbar spine stability. Journal of Biomechanics 33: 1377-1385

8: 388--395

and/or increased intra-abdominal pressure. European Spine Journal

Lumbar spine stability can be augmented with an abdominal belt



Cholewicki J, Ivancic P C, Radebold A 2002a Can increased intra­ abdominal pressure in humans be decoupled from trunk muscle co­ contraction during steady state isometric exertions? European Journal of Applied Physiology 87: 1 27-133 Cholewicki J, Polzhofer G K, Galloway M T, Greene H S, Shah R A, Radebold A 2002b Neuromuscular function in athletes following recovery from an acute low back injury. Journal of Orthopaedic and Sports Physical Therapy 32: 569-576 Ciriello V M, Snook S H 1995 The effect of back belts on lumbar muscle fatigue. Spine 20: 1271-1 278 Cobb C R, deVries H A, Urban R T, Luekens C A, Bagg R J 1975 Electrical activity in muscle pain. American Journal of Physical Medicine 54: 80-87 Cresswell A G, Thorstensson A 1989 The role of the abdominal musculature in the elevation of the intra-abdominal pressure during specified tasks. Ergonomics 32: 1237-1246 Cresswell A G, Thorstensson A 1994 Changes in intra-abdominal pressure, trunk muscle activation and force during isokinetic lifting and lowering. European Journal of Applied Physiology and Occupational Physiology 68: 315-321 Cresswell A G, Grundstrom H, Thorstensson A 1992 Observations on intra-abdominal pressure and patterns of abdominal intra-muscular activity in man. Acta Physiologica Scandinavica 144: 409-418 Cresswell A G, Oddsson L, Thorstensson A 1994 The influence of sudden perturbations on trunk muscle activity and intra-abdominal pressure while standing. Experimental Brain Research 98: 336-341 Crisco J J 3rd, Panjabi M M 1991 The intersegmental and multisegmental muscles of the lumbar spine: a biomechanical model comparing lateral stabilizing potential. Spine 16: 793-799 Crisco J J, Panjabi M M, Yamamoto I, Oxland T R 1992 Euler stability of the human ligamentous lumbar spine. II: Experiment. Clinical Biomechanics 7: 27-32 Cyron B M, Hutton W C, Stott J R 1979 Spondylolysis: the shearing stiffness of the lumbar intervertebral joint. Acta Orthopaedica Belgica 45: 459-469 Daggfeldt K, Thorstensson A 1997 The role of intra-abdominal pressure in spinal unloading. Journal of Biomechanics 30: 1149-1155 Damkot D K, Pope M H, Lord J, Frymoyer J W 1984 The relationship between work history, work environment and low-back pain in men. Spine 9: 395-399 Davies G J, Dickoff-Hoffrnan S 1993 Neuromuscular testing and rehabilitation of the shoulder complex. Journal of Orthopaedic and Sports Physical Therapy 18: 449-458 Davis K G, Marras W S, Walters T R 1998 Evaluation of spinal loading during lowering and lifting. Clinical Biomechanics 13: 141-152 De Luca D J 1993 Use of the surface EMG signal for performance evaluation of back muscles. Nerve Muscle 16: 210-216 Deyo R A, Weinstein J N 2001 Primary care: low back pain. New England Journal of Medicine 344: 363-370 Diener H C, Bootz F, Dichgans J, Bruzek W 1983 Variability of postural 'reflexes' in humans. Experimental Brain Research 52: 423-428 Edgerton V R, Wolf S L, Levendowski D J, Roy R R 1996 Theoretical basis for patterning EMG amplitudes to assess muscle dysfunction. Medicine and Science in Sports and Exercise 28: 744-751 Eils E, Rosenbaum D 2001 A multi-station proprioceptive exercise program. in patients with ankle instability. Medicine and Science in Sports and Exercise 33: 1991-1998 Fairbank J C, O'Brien J P, Davis P R 1980 Intraabdominal pressure rise during weight lifting as an objective measure of low-back pain. Spine 5: 1 79-184 Feyer A M, Herbison P, Williamson A M et al 2000 The role of physical and psychological factors in occupational low back pain: a prospective cohort study. Occupational and Environmental Medicine 57: 116-120 Fidler M W, Plasmans C M 1983 The effect of four types of support on the segmental mobility of the lumbosacral spine. Journal of Bone and Joint Surgery (American volume) 65: 943-947

Fitzgerald G 1998 Non-operative anterior cruciate ligament rehabilitation for individuals participating in high level physical activity. Rehabilitation Sciences 165. Allegheny University of the Health Sciences, Philadelphia Frank J W, Kerr M S, Brooker A S et al 1996 Disability resulting from occupational low back pain. I: What do we know about primary prevention? A review of the scientific evidence on prevention before disability begins. Spine 21 : 2908-291 7 Fritz J M , Erhard R E, Hagen B F 1998 Segmental instability o f the lumbar spine. Physical Therapy 78: 889-896 Fryrnoyer J W, Pope M H, Clements J H, Wilder D G, MacPherson B, Ashikaga T 1983 Risk factors in low-back pain: an epidemiological survey. Journal of Bone and Joint Surgery (American volume) 65: 213-218 Gardner-Morse M G, Stokes I A 1998 The effects of abdominal muscle co-activation on lumbar spine stability. Spine 23: 86-91 Gardner-Morse M, Stokes I A, Laible J P 1995 Role of muscles in lumbar spine stability in maximum extension efforts. Journal of Orthopaedic Research 13: 802-808 Gielen C C, Houk J C 1987 A model of the motor servo: incorporating nonlinear spindle receptor and muscle mechanical properties. Biological Cybernetics 57: 21 7-231 Gill K P, Callaghan M J 1998 The measurement of lumbar proprioception in individuals with and without low back pain. Spine 23: 371-377 Gracovetsky S, Farfan H, Helleur C 1985 The abdominal mechanism. Spine 10: 317-324 Gracovetsky S, Kary M, Pitchen I, Levy S, Ben Said R 1989 The importance of pelvic tilt in reducing compressive stress in the spine during flexion---€xtension exercises. Spine 14: 412-416 Gracovetsky S, Kary M, Levy S, Ben Said R, Pitchen I, Helie J 1990 Analysis of spinal and muscular activity during flexion /extension and free lifts. Spine 15: 1333-1339 Granata K P, Marras W S 1995 An EMG-assisted model of trunk loading during free-dynamic lifting. Journal of Biomechanics 28: 1309-1317 Granata K P, Marras W S 2000 Cost-benefit of muscle cocontraction in protecting against spinal instability. Spine 25: 1398-1404 Granata K P, Orishimo K F 2001 Response of trunk muscle co­ activation to changes in spinal stability. Journal of Biomechanics 34: 1 1 1 7-1123 Granata K P, Marras W S, Davis K G 1997 Biomechanical assessment of lifting dynamics, muscle activity and spinal loads while using three different styles of lifting belt. Clinical Biomechanics 12: 107-115 Granhed H, Jonson R, Hansson T 1989 Mineral content and strength of lumbar vertebrae: a cadaver study. Acta Orthopaedica Scandinavica 60: 105-109 Greene H S, Cholewicki J, Galloway M T, Nguyen C V, Radebold A 2001 A history of low back injury is a risk factor for recurrent back injuries in varsity athletes. American Journal of Sports Medicine 29: 795-800 Grenier N, Greselle J F, Vital J M et al 1989 Normal and disrupted lumbar longitudinal ligaments: correlative MR and anatomic study. Radiology 171: 197-205 Hamaoui A, Do M, Poupard L, Bouisset S 2002 Does respiration perturb body balance more in chronic low back pain subjects than in healthy subjects? Clinical Biomechanics (Bristol, Avon) 17: 548 Harman E A, Rosenstein R M, Frykman P N, Nigro G A 1989 Effects of a belt on intra-abdominal pressure during weight lifting. Medicine and Science in Sports and Exercise 2 1 : 186-190 Hatta I, Sugi H, Tamura Y 1988 Stiffness changes in frog skeletal muscle during contraction recorded using ultrasonic waves. Journal-of Physiology 403: 193-209 Hemborg B, Moritz U 1985 Intra-abdominal pressure and trunk muscle Journal of Rehabilitation Medicine 17: 5-13 activity during lifting. II: Chronic low-back patients. Scandinavian -

C l i n i c a l b i o m ec h a n ics of the l u m b a r s p i n e


Hides J A, Jull G A, Richardson C A 2001 Long-term effects of specific stabilizing exercises for first-episode low back pain. Spine 26: E243:-248 Hodges P W 2001 Changes in motor planning of feed-forward postural responses of the trunk muscles in low back pain. Experimental Brain Research 1 41 : 261-266 Hodges P W, Richardson C A 1996 Inefficient muscular stabilization of the lumbar spine associated with low back pain: a motor control evaluation of transversus abdominis. Spine 2 1 : 2640-2650 Hodges P W, Richardson C A 1997a Feed-forward contraction of transversus abdominis is not influenced by the direction of arm movement. Experimental Brain Research 114: 362-370 Hodges P W, Richardson C A 1997b Contraction of the abdominal muscles associated with movement of the lower limb. Physical Therapy 77: 132-142; discussion 142-134 Hodges P W, Richardson C A 1998 Delayed postural contraction of transversus abdominis in low back pain associated with movement of the lower limb. Journal of Spinal Disorders 11: 46-56 Hodges P W, Richardson C A 1999 Altered trunk muscle recruitment in people with low back pain with upper limb movement at different speeds. Archives of PhYSical Medicine and Rehabilitation 80: 1005-1012 Hodges P Cresswell A, Thorstensson A 1999 Preparatory trunk motion , accompanies rapid upper limb movement. Experimental Brain Research �24: 69-79 Hodges P W, Cresswell A G, Daggfeldt K, Thorstensson A 2001 In vivo measurement of the effect of intra-abdominal pressure on the human spine. Journal of Biomechanics 34: 347-353 Hogan N 1990 Mechanical impedance of single- and multi-articular systems. In: Winters J M, Woo S L (eds) Multiple muscle systems: biomechanics and movement organization. Springer-Verlag, New York, pp 149-164. Holmstrom E, Moritz U 1992 Effects of lumbar belts on trunk muscle strength and endurance: a follow-up study of construction workers. Journal of Spinal Disorders 5: 260-266 Huang Q M, Hodges P W, Thorstensson A 2001 Postural control of the

Jull G , Richardson, C A , Toppenberg R, Comerford M , Bui B 1 993 Towards a measurement of active muscle control for lumbar stabilisation. Australian Physiotherapy 39: 187-193 Keith A 1923 Man's posture: its evolution and disorders. Lecture IV: The adaptations of the abdomen and its viscera to the orthograde posture. British Medical Journal I: 587-590 Kelsey J L, Githens P B, White A A D et al 1 984 An epidemiologic study of lifting and twisting on the job and risk for acute prolapsed lumbar intervertebral disc. Journal of Orthopaedic Research 2: 61-66 Kerr M S, Frank J W, Shannon H S et al 2001 Biomechanical and psychosocial risk factors for low back pain at work. American Journal of Public Health 9 1 : 1 069-1075 Kingma I, de Looze M P, van Dieen J H, Toussaint H M, Adams M A, Baten C T 1 998 When is a lifting movement too asymmetric to identify low-back loading by 2-D analysis? Ergonomics 41: 1453-1461 Kumar S 2001 Theories of musculoskeletal injury causation. Ergonomics 44: 1 7-47 Lander J E, Simonton R L, Giacobbe J K 1990 The effectiveness of weight-belts during the squat exercise. Medicine and Science in Sports and Exercise 22: 1 1 7-126 Lander J E, Hundley J R, Simonton R L 1992 The effectiveness of weight-belts during multiple repetitions of the squat exercise. Medicine and Science in Sports and Exercise 24: 603-609 Lantz S A, Schultz A B 1986a Lumbar spine orthosis wearing. II: Effect on trunk muscle myoelectric activity. Spine 1 1 : 838-842 Lantz S A, Schultz A B 1986b Lumbar spine orthosis wearing. I : Restriction o f gross body motions. Spine 1 1 : 834-837 Lariviere C, Gagnon D, Loisel P 2000 The comparison of trunk muscles EMG activation between subjects with and without chronic low back pain during flexion--extension and lateral bending tasks. Journal of Electromyography and Kinesiology 10: 79-91 Latash M L, Scholz J P, Schoner G 2002 Motor control strategies revealed in the structure of motor variability. Exercise and Sport Sciences Reviews 30: 26-31 Lavender S A, Tsuang Y H, Andersson G B, Hafezi A, Shin C C 1992a Trunk muscle cocontraction: the effects of moment direction and moment magnitude. Journal of Orthopaedic Research 10: 691-700 Lavender S A, Tsuang Y H, Hafezi A, Andersson G B, Chaffin D B, Hughes R E 1992b Co-activation of the trunk muscles during asymmetric loading of the torso. Human Factors 34: 239-247 Lavender S A, Tsuang Y H, Andersson G B 1993 Trunk muscle activation and cocontraction while resisting applied moments in a twisted posture. Ergonomics 36: 1145-1157 Lee Y H, Kang S M 2002 Effect of belt pressure and breath held on

trunk in response to lateral support surface translations during trunk movement and loading. Experimental Brain Research 141:
552-559 Hughes R E, Bean J C, Chaffin D B 1 995 Evaluating the effect of co-contraction in optimization models. Journal of Biomechanics 28: 875-878 Hughes R E, Bean J C, Chaffin D B 2001 A method for classifying co-contraction of lumbar muscle activity. Journal of Applied Biomechanics 17: 253:-258 Hutton W C, Cyron B M, Stott J R 1979 The compressive strength of lumbar vertebrae. Journal of Anatomy 129: 753:-758 Ivancic P C, Cholewicki J, Radebold A 2002 Effects of the abdominal belt on muscle-generated spinal stability and L4/15 joint compression force. Ergonomics 45: 501-513 Jackson M, Solomonow M, Zhou B, Baratta R V, Harris M 2001 Multifidus EMG and tension-relaxation recovery after prolonged static lumbar flexion. Spine 26: 715-723 Janevic J, Ashton-Miller J A, Schultz A B 1991 Large compressive preloads decrease lumbar motion segment flexibility. Journal of Orthopaedic Research 9: 228-236 Jellema P, van Tulder M W, van Poppel M N, Nachemson A L, Bouter L M 2001 Lumbar supports for prevention and treatment of low back pain: a systematic review within the framework of the Cochrane Back Review Group. Spine 26: 377-386 Jonsson B 1978 KineSiology: with special reference to electromyographic kinesiology. Electroencephalography and Clinical Neurophysiology 34 (Supp!.): 417-428 Jull G A, Richardson C A 2000 Motor control problems in patients with spinal pain: a new direction for therapeutic exercise. Journal of Manipulative and Physiological Therapeutics 23: 115-117

trunk electromyography. Spine 27: 282-290
Leipholz H H E 1987 Stability theory: an introduction to the stability of dynamic systems and rigid bodies. John Wiley, Stuttgart Lucas D B, Bresler B 1961 Stability of the ligamentous spine. Biomechanics Laboratory, University of California, San Francisco Lund J P, Donga R, Widmer C G, Stohler C S 1991 The pain-adaptation model: a discussion of the relationship between chronic musculoskeletal pain and motor activity. Canadian Journal of Physiology and Pharmacology 69: 683-694 Luoto S, Taimela S, Hurri H, Aalto H, Pyykko I, Alaranta H 1996 Psychomotor speed and postural control in chronic low back pain patients: a controlled follow-up study. Spine 2 1 : 2621-2627 Luoto S, Taimela S, Hurri H, Alaranta H 1999 Mechanisms explaining the association between low back trouble and deficits in information processing: a controlled study with follow-up. Spine 24: 255-261 McCowin P R, Borenstein D, Wiesel S W 1991 The current approach to the medical diagnosiS of low back pain. Orthopedic Clinics of North America 22: 315-325 McGill S M 1 992a A myoelectrically based dynamic three-dimensional model to predict loads on lumbar spine tissues during lateral bending. Journal of Biomechanics 25: 395-414



McGill S M B S 1992b Creep response of the lumbar spine to prolonged full flexion. Clinical Biomechanics 7: 43--48 McGill S M 1998 Low back exercises: evidence for improving exercise regimens. Physical Therapy 78: 754-765 McGill S M 1997 The biomechanics of low back injury: implications on current practice in industry and the clinic. Journal of Biomechanics 30: 465-475 McGill S M, Norman R W 1987 Reassessment of the role of intra­ abdominal pressure in spinal compression. Ergonomics 30: 1565-1588 McGill S M, Norman R W 1988 Potential of lumbodorsal fascia forces to generate back extension moments during squat lifts [see comments). Journal of Biomedical Engineering 10: 312-318 McGill S M, Sharratt M T 1990 Relationship between intra-abdominal pressure and trunk EMG. Clinical Biomechanics 5: 59--{)7 McGill S M, Norman R W, Sharratt M T 1990 The effect of an abdominal belt on trunk muscle activity and intra-abdominal pressure during squat lifts. Ergonomics 33: 147-160 McGill S, Seguin 1, Bennett G 1994 Passive stiffness of the lumbar torso in flexion, extension, lateral bending, and axial rotation: effect of belt wearing and breath holding. Spine 19: 696-704 McGill S M, Sharratt M T, Seguin J P 1995 Loads on spinal tissues during simultaneous lifting and ventilatory challenge. ErgonOmics 38: 1772-1792 McGill S M, Norman R W, Cholewicki J 1996 A simple polynomial that predicts low-back compression during complex 3-D tasks. Ergonomics 39: 1107-1118 McNair P J, Heine P J 1999 Trunk proprioception: enhancement through lumbar bracing. Archives of Physical Medicine and Rehabilitation 80: 96-99 Magnusson M L, Aleksiev A, Wilder 0 G et al 1996 Unexpected load and asymmetric posture as etiologic factors in low back pain. European Spine Journal 5: 23-35 Maitland M E, Ajemian S V, Suter E 1999 Quadriceps femoris and hamstring muscle function in a person with an unstable knee. Physical Therapy 79: 66-75 Majkowski G R, Jovag B W, Taylor B T et al 1998 The effect of back belt use on isometric lifting force and fatigue of the lumbar paraspinal muscles. Spine 23: 2104-2109 Manning 0 P, Mitchell R G, Blanchfield L P 1984 Body movements and events contributing to accidental and nonaccidental back injuries. Spine 9: 734-739 Mannion A F, Dolan P 1994 Electromyographic median frequency changes during isometric contraction of the back extensors to fatigue. Spine 19: 1223-1229 Marras W S, Mirka G A 1996 Intra-abdominal pressure during trunk extension motions. Clinical Biomechanics 11: 267-274 Marras W S, Lavender S A, Leurgans S E et al 1995 Biomechanical risk factors for occupationally related low back disorders. Ergonomics 38: 377-410 Marras W S, Jorgensen M 1, Davis K G 2000 Effect of foot movement and an elastic lumbar back support on spinal loading during free­ dynamic symmetric and asymmetric lifting exertions. Ergonomics 43: 653--{)68 Marras W S, Davis K G, Ferguson S A, Lucas B R, Gupta P 2001 Spine loading characteristics of patients with low back pain compared with asymptomatic individuals. Spine 26: 2566-2574 Mientjes M I, Frank J S 1999 Balance in chronic low back pain patients compared to healthy people under various conditions in upright standing. Clinical Biomechanics (Bristol, Avon) 14: 710-716 Miller J A, Schultz A B, Warwick 0 N, Spencer 0 L 1986 Mechanical properties of lumbar spine motion segments under large loads. Journal of Biomechanics 19: 79-84 Million R, Nilsen K H, Jayson M I, Baker R 0 1981 Evaluation of low back pain and assessment of lumbar corsets with and without back supports. Annals of the Rheumatic Diseases 40: 449-454

Morgan 0 L 1977 Separation of active and passive components of short-range stiffness of muscle. American Journal of Physiology 232: C45-49 Morris J M, Lucas 0 B, Bresler B 1961 The role of trunk in stability of the spine. Journal of Bone and Joint Surgery (American volume) 43: 327-351 Nachemson A L, Andersson B 1, Schultz A B 1986 Valsalva maneuver biomechanics: effects on lumbar trunk loads of elevated intra abdominal pressures. Spine 11: 476-479 Narloch J A, Brandstater M E 1995 Influence of breathing technique on arterial blood pressure during heavy weight lifting. Archives of Physical Medicine and Rehabilitation 76: 457-462 Nashner L M, Cordo P J 1981 Relation of automatic postural responses and reaction-time voluntary movements of human leg muscles. Experimental Brain Research 43: 395-405 Newcomer K, Laskowski E R, Yu B, Johnson J C, An K N 2001 The effects of a lumbar support on repositioning error in subjects with low back pain. Archives of Physical Medicine and Rehabilitation 82: 906-910 Norman R, Wells R, Neumann P et al 1998 A comparison of peak vs cumulative physical work exposure risk factors for the reporting of low back pain in the automotive industry. Clinical Biomechanics 13: 561-573 Norris C 1995 Spinal stabilization. 5: An exercise programme to enhance lumbar stabilization. Physiotherapy 81: 138-146 Oddsson L I, Persson T, Cresswell A G, Thorstensson A 1999 Interaction between voluntary and postural motor commands during perturbed lifting. Spine 24: 545-552 Ohana N, Sheinis 0, Rath E, Sasson A, Atar 0 2000 Is there a need for lumbar orthosis in mild compression fractures of the thoracolumbar spine? A retrospective study comparing the radiographic results between early ambulation with and without lumbar orthosis. Journal of Spinal Disorders 13: 305-308 amino K, Hayashi Y 1992 Preparation of dynamic posture and occurrence of low back pain. Ergonomics 35: 693-707 O'Sullivan P B 2000 Lumbar segmental 'instability': clinical presentation and specific stabilizing exercise management. Manual Therapy 5: 2-12 O'Sullivan P, Twomey L, Allison G, Sinclair J, Miller K 1997a Altered patterns of abdominal muscle activation in patients with chronic low back pain. Australian Journal of Physiotherapy 43: 91-98 O'Sullivan P B, Twomey L T, Allison G T 1997b Evaluation of specific stabilizing exercise in the treatment of chronic low back pain with radiologic diagnosis of spondylolysis or spondylolisthesis. Spine 22: 2959-2967 O'Sullivan P B, Twomey L, Allison G T 1998 Altered abdominal muscle recruitment in patients with chronic back pain following a specific exercise intervention. Journal of Orthopaedic and Sports Physical Therapy 27: 114-124 Osvalder A L, Neumann P, Lovsund P, Nordwall A 1990 Ultimate strength of the lumbar spine in flexion - an in vitro study. Journal of Biomechanics 23: 453--460 Osvalder A L, Neumann P, Lovsund P, Nordwall A 1993 A method for studying the biomechanical load response of the (in vitro) lumbar spine under dynamic flexion-shear loads. Journal of Biomechanics 26: 1227-1236 Oxland T R, Panjabi M M 1992 The onset and progression of spinal injury: a demonstration of neutral zone sensitivity. Journal of Biomechanics 25: 1165-1172 Panjabi M M 1992a The stabilizing system of the spine. I: Function, dysfunction, adaptation, and enhancement. Journal of Spinal Disorders 5: 383-389 Panjabi M M 1992b The stabilizing system of the spine. II: Neutral zone and instability hypothesis. Journal of Spinal Disorders 5: 390-396 Paquet N, Malouin F, Richards C L 1994 Hip-spine movement interaction and muscle activation patterns during sagittal trunk movements in low back pain patients. Spine 19: 596--{)03 -

C l i n i c a l b i o m e c h a n ics of t h e l u m ba r s p i n e


Parkhurst T M, Burnett C N 1994 Injury and proprioception in the 19: 282-295

lower back. Journal of Orthopaedic and Sports Physical Therapy

ligament injury: a comparison o f patients who compensate well for Electromyography and KineSiology 8: 349-362 the injury and those who require operative stabilization. Journal of

Patwardhan A G, Havey R M, Meade K P, Lee B, Dunlap B 1999 A spine in compression. Spine 24: 1003-1009

follower load increases the load-carrying capacity of the lumbar

Saal J A, Saal J S 1989 Nonoperative treatment of herniated lumbar 14: 431-437

intervertebral disc with radiculopathy: an outcome study. Spine

Peach J P, Sutarno C G, McGill S M 1998 Three-dimensional kinematics and trunk muscle myoelectriC activity in the young lumbar spine: a database. Archives of Physical Medicine and Rehabilitation 79: 663-669

Saal J A, Saal J S, Herzog R J 1990 The natural history of lumbar 683-686

intervertebral disc extrusions treated nonoperatively. Spine 15:

Pope M H, Panjabi M 1985 Biomechanical definitions of spinal Pope M H, Magnusson M, Wilder D G 1998 Low back pain and whole body vibration. Clinical Orthopaedics and Related Research 354: 241-248 instability. Spine 10: 255-256

Schneider G A, Bigelow C, Amoroso P J 2000 Evaluating risk of re­ injury among 1214 army airborne soldiers using a stratified 156-163 survival model. American Journal of Preventive Medicine 18:

Schultz A, Haderspeck K, Warwick D, Portillo D 1983 Use of complex standing tasks. Journal of Orthopaedic Research 1: 77-91

Porter R W, Adams M A, Hutton W C 1989 Physical activity and the Porth C J, Bamrah V S, Tristani F E, Smith J J 1984 The Valsalva 13: 507-518 strength of the lumbar spine. Spine 14: 201-203

lumbar trunk muscles in isometric performance of mechanically

maneuver: mechanisms and clinical implications. Heart and Lung

Schwarzer A C, Aprill C N, Derby R, Fortin J, Kine G, Bogduk N 1995 The prevalence and clinical features of internal disc disruption in patients with chronic low back pain. Spine 20: 1878-1883

Potvin J R 1997 Use of NlOSH equation inputs to calculate lumbosacral Potvin J R, Norman R W, McGill S M 1991 Reduction in anterior shear forces on the L4/LS disc by the lumbar musculature. Clinical Biomechanics 6: 88-96 compression forces. Ergonomics 40: 691-707

Sihvonen T, Partanen J, Hanninen 0, Soimakallio S 1991 Electric back pain patients and healthy controls. Archives of Physical Medicine and Rehabilitation 72: 1080-1087

behavior of low back muscles during lumbar pelvic rhythm in low

Sihvonen T, Lindgren K A, Airaksinen 0, Manninen H 1997 Movement electromyographic findings in recurrent low back pain. Spine disturbances of the lumbar spine and abnormal back muscle

Rabinowitz D, Bridger R S, Lambert M 1 1998 Lifting technique and investigation. Safety Science 28: 155-164

abdominal belt usage: a biomechanical, physiological and subjective

Rack P M, Westbury D R 1973 The short range stiffness of active of Physiology 229: 16P-I7P

Silfies S 2002 Trunk muscle and motor control impairments in patients with lumbar instability. Department of Rehabilitation Sciences 184. Drexel University, Philadelphia

22: 289-295

mammalian muscle and its effect on mechanical properties. Journal

Rack P M, Westbury D R 1974 The short range stiffness of active of Physiology 240: 331-350

mammalian muscle and its effect on mechanical properties. Journal

Smith E B, Rasmussen A A, Lechner D E, Gossman M R, Quintana J B, Grubbs B L 1996 The effects of lumbosacral support belts and women. Spine 21: 356-366 abdominal muscle strength on functional lifting ability in healthy

Radebold A, Cholewicki J, Panjabi M M, Patel T C 2000 Muscle

Radebold A, Cholewicki J, Polzhofer G K, Greene H S 2001 Impaired postural control of the lumbar spine is associated with delayed pain. Spine 26: 724-730 muscle response times in patients with chronic idiopathic low back

and in patients with chronic low back pain. Spine 25: 947-954

response pattern to sudden trunk loading in healthy individuals

Solomonow M, Eversull E, He Zhou B, Baratta R V, Zhu M P 2001 during cyclic lumbar flexion. Spine 26: E314-324

Neuromuscular neutral zones associated with viscoelastic hysteresis

Solomonow M, Zhou B, Baratta R V, Zhu M, Lu Y 2002 Neuromuscular Journal of Electromyography and Kinesiology 12: 81-90 disorders associated with static lumbar flexion: a feline model.

Ramos C F, Hacisalihzade S S, Stark L W 1990 Behaviour space of a voluntary movement. Medical and Biological Engineering and

stretch reflex model and its implications for the neural control of

Stein R B, Gordon T 1986 Nonlinear stiffness - force relationships in and Pharmacology 64: 1236-1244

whole mammalian skeletal muscles. Canadian Journal of Physiology

Reddell C R, Congleton J J, Huchingson R D, Montgomery J F 1992 An evaluation of a weightlifting belt and back injury prevention 23: 319-329 training class for airline baggage handlers. Applied Ergonomics

Computing 28: 15-23

Stein R B, Oguztoreli M N 1984 Modification of muscle responses by Sterling M, Jull G, Wright A 2001 The effect of musculoskeletal pain on Stokes I, Gardner-Morse M 1999 Lumbar spinal muscle activation motor activity and control. Journal of Pain 2: 135-145 spinal circuitry. Neuroscience 1 1 : 231-240

Reyna J R J, Leggett S H, Kenney K, Holmes B, Mooney V 1995 The Richardson C A, Jull G A 2000 Muscle control-pain control: what Rissanen P M 1960 The surgical anatomy and pathology of the special reference to ligament ruptures. Acta Orthopaedica Scandinavica 46(Suppl.): 1-100 exercises would you prescribe? Manual Therapy 1: 2-10

effect of lumbar belts on isolated lumbar muscle. Spine 20: 68-73

synergies predicted by multi-criteria cost function. Proceedings of 45th Annual Meeting, Orthopaedic Research Society, Anaheim, California

Stokes I A, Gardner-Morse M, Churchill D, Laible J P 2002 of Biomechanics 35: 517-521

supraspinous and interspinous ligaments of the lumbar spine, with

Measurement of a spinal motion segment stiffness matrix. Journal

Sullivan M S, Mayhew T P 1995 The effect of lumbar support belts on Occupational Rehabilitation 5: 131-143 isometric force production during a simulated lift. Journal of

Rohlmann A, Graichen F, Weber U, Bergmann G 2000 Monitoring in vivo implant loads with a telemeterized internal spinal fixation device. Spine 25: 2981-2986

Taimela S, Osterman K, Alaranta H, Soukka A, Kujala U M 1993 Long preliminary report. Archives of PhYSical Medicine and Rehabilitation 74: 1161-1164 psychomotor reaction time in patients with chronic low-back pain:

Roland M 0 1986 A critical review of the evidence for a pain-.spasm-pain Rozzi S L, Lephart S M, Sterner R, Kuligowski L 1999 Balance training for persons with functionally unstable ankles. Journal of Orthopaedic and Sports Physical Therapy 29: 478-486 cycle in spinal disorders. Clinical Biomechanics 1 : 102-109

Taimela S, Kankaanpaa M, Luoto S 1999 The effect of lumbar fatigue on the ability to sense a change in lumbar position: a controlled study. Spine 24: 1322-1327

Rudolph K S, Eastlack M E, Axe M J, Snyder-Mackler L 1998 Basmajian Student Award paper. Movement patterns after anterior cruciate

Takala E-P, Korhonen I, V ilkari-Juntura E 1997 Postural sway and

stepping response among working population: reproducibility,



long-term stability, and associations with symptoms of the low back. Clinical Biomechanics 12: 429-437 Tropp H, Ekstrand J, Gillquist J 1984 Stabilometry in functional instability of the ankle and its value in predicting injury. Medicine and Science in Sports and Exercise 16: 64-66 Troup J D, Martin J W, Lloyd D C 1981 Back pain in industry: a prospective survey. Spine 6: 61-69 Tuong N H, Dansereau J, Maurais G, Herrera R 1998 Three­ dimensional evaluation of lumbar orthosis effects on spinal behavior. Journal of Rehabilitation Research and Development 35: 34-42 van Dieen J H, de Looze M P 1999a Directionality of anticipatory activation of trunk muscles in a lifting task depends on load knowledge. Experimental Brain Research 128: 397-404 van Dieen J H, de Looze M P 1999b Sensitivity of single-equivalent trunk extensor muscle models to anatomical and functional assumptions. Journal of Biomechanics 32: 195-198 van Dieen J H, Weinans H, Toussaint H M 1999 Fractures of the lumbar vertebral endplate in the etiology of low back pain: a hypothesis on the causative role of spinal compression in a specific low back pain. Medical Hypotheses 53: 246-252 van Dieen J H, Dekkers J J, Groen V, Toussaint H M, Meijer 0 G 2001 Within-subject variability in low back load in a repetitively performed, mildly constrained lifting task. Spine 26: 1 799-1804 van Dieen J H, Cholewicki J, Radebold A 2003 Trunk muscle recruitment patterns in low back pain patients enhance the stability of the lumbar spine. Spine 28: 834-841 Van Dillen L R, Sahrmann S A, Norton B J et al 1998 Reliability of physical examination items used for classification of patients with low back pain. Physical Therapy 78: 979-988 Van Dillen L R, Sahrmann S A, Norton B J et al 2001 Effect of active limb movements on symptoms in patients with low back pain. Journal of Orthopaedic and Sports Physical Therapy 31: 402-413; discussion 414-408 van Poppel M N, de Looze M P, Koes B W, Smid T, Bouter L M 2000 Mechanisms of action of lumbar supports: a systematic review. Vera-Garcia F j, Grenier S G, McGill S M 2000 Abdominal muscle response during curl-ups on both stable and labile surfaces. Physical Therapy 80: 564-569 Vereijken B, van Emmerik R, Whiting H, Newell K 1992a Free(z)ing degrees of freedom in skill acquisition. Journal of Motor Behavior 24: 133-142 Vereijken B, Whiting H T, Beek W J 1992b A dynamical systems approach to skill acquisition. Quarterly Journal of Experimental Psychology 45 A: 323-344 Walsh N E, Schwartz R K 1990 The influence of prophylactic orthoses on abdominal strength and low back injury in the workplace. American Journal of PhYSical Medicine and Rehabilitation 69: 245-250 Spine 25: 2103-2113

Warren L P, Appling S, Oladehin A, Griffin J 2001 Effect of soft lumbar support belt on abdominal oblique muscle activity in nonimpaired adults during squat lifting. Journal of Orthopaedic and Sports Physical Therapy 3 1 : 316-323 Wassell J T, Gardner L I, Landsittel D P, Johnston J J, Johnston J M 2000 A prospective study of back belts for prevention of back pain and injury. JAMA 284: 2727-2732 White ill A, Panjabi M 1990 Clinical instability. In: White ill A, Panjabi M (eds) Clinical biomechanics of the spine. Lippincott, Philadelphia, pp 342-360 Wilder D G, Aleksiev A R, Magnusson M L, Pope M H, Spratt K F, Goel V K 1996 Muscular response to sudden load: a tool to evaluate fatigue and rehabilitation. Spine 21: 2628-2639 Wilk K E, Meister K, Andrews J R 2002 Current concepts in the rehabilitation of the overhead throwing athlete. American Journal of Sports Medicine 30: 136-151 Winters J M 1995 An improved muscle-reflex actuator for use in large­ scale neuro-musculoskeletal models. Annals of Biomedical Engineering 23: 359-374 Winters J, Stark L, Seif-Naraghi A H 1988 An analysis of the sources of musculoskeletal system impedance. Journal of Biomechanics 21 : 1011-1025 Woldstad J C, Sherman B R 1998 The effects of a back belt on posture, strength, and spinal compressive force during static lift exertions. International Journal of Industrial Ergonomics 22: 409-416 Yettram A L, Jackman M J 1980 Equilibrium analysis for the forces in the human spinal column and its musculature. Spine 5: 402-411 Zahalak G I 1986 A comparison of the mechanical behavior of the cat soleus muscle with a distribution-moment model. Journal of Biomechanical Engineering 108: 131-140 Zahalak G I, Heyman S J 1979 A quantitative evaluation of frequency response characteristics of active human skeletal muscle in vivo. Journal of Biomechanical Engineering 101: 28-37 Zahalak G I, Ma S P 1990 Muscle activation and contraction: constitutive relations based directly on cross-bridge kinetics. Journal of Biomechanical Engineering 112: 52-62 Zajac F E, Winters J M 1990 Modeling musculoskeletal movement systems: joint and body segmental dynamiCS, musculoskeletal actuation, and neuromuscular control. In: Winters J M, Woo S L-Y (eds) Multiple muscle systems: biomechanics and movement organization: Springer-Verlag, New York, pp 121-148. Zattara M, Bouisset S 1988 Posturo-kinetic organisation during the early phase of voluntary upper limb movement. 1 : Normal subjects. Journal of Neurology, Neurosurgery and Psychiatry 5 1 : 956-965 Zedka M, Prochazka A, Knight B, Gillard D, Gauthier M 1999 Voluntary and reflex control of human back muscles during induced pain. Journal of PhYSiology 520 (2): 591-604




' Clinical biomechanics of lifting
s. Mi\anese




Characteristics of lifting tasks


Approaches used t o study the effect of lifting

Physiological approach 90 Psychophysical approach 90 Biomechanical approach 90 Integrated approaches 9 1
Biomechanics of lifting



Biomechanical criteria for determining safe lifting 92 Style of lifting 93 Squat lifting 93 Stoop lifting 93 Semi-squat lifting 93 Spinal motion segment 94 Intradiscal pressure 96 Muscle activity 96 Intra-abdominal pressure 97 Other joints 98
Risk factors

Personal risk factors 98 Environmental personal risk factors Job related risk factors 99 Horizontal position of the load 99 Vertical position of the load 1 00 Lifting frequency 1 00 Use of handles 1 01 Weight of the load 1 01 Asymmetry of the lift 1 01




Despite the increasing use of risk management pro­ grammes in industry, musculoskeletal injuries attributed to manual handling remain a major burden to the community in terms of financial costs and human suffering (Waters & Putz Anderson 1996, Chaffin et al 1999), Mechanical low back pain in particular remains a major health and safety issue for both the clinic based and the industrial physio­ therapist alike. The use of mechanization and ergonomic re­ engineering in the production process and the popularity of manual handling training programmes for workers appears to have done little to reduce the prevalence of low back pain. Given the increasing role of physiotherapists in the design and implementation of manual handling risk management programmes, it is pertinent for us to revisit the scientific basis underpinning our understanding of the risks involved in manual handling. A review of the epidemiological literature on low back pain (Hildebrandt 1987) found 24 work-related factors reported by at least one published source as being associ­ ated with low back pain. These factors reflected those of an earlier landmark review that identified that heavy physical loading, manual handling, including lifting, bending, twist­ ing, sitting, sustained non-neutral postures and vehicular driving, were associated with an increased risk of low back pain (Magora 1973). Chaffin & Park reported that workers involved in heavy lifting were at least eight times more likely to report suffering back injuries as those workers per­ forming sedentary work tasks (Chaffin & Park 1973). Despite the published evidence, the role of occupational risk factors in the development of disc degeneration and low back pain remains controversial. It has been reported that familial aggregation, age and other unexplained fac­ tors might play a more important role in disc degeneration than occupational loading factors (Videman & Battie 1999). It would appear to be prudent to conclude at this stage, in the absence of conclusive evidence, that the causes for low back pain are multifactorial, as indeed are the optimal man­ agement approaches.




Psychophysical approach
Psychophysics examines the relationship between the per­ ception of human sensations and physical stimuli (Waters & Putz Anderson 1996). Proponents of this approach believe that the worker 's actual level of workload can be assessed by his/her subjective judgement or perception of physical stress (Waters & Putz Anderson 1996). Typical studies involve the measurement of maximal acceptable weight limits (MAWL) for specific task conditions and for various workers. The results of such measures allow the generation of tables of acceptable weight limits for various segments of the population (Snook & Ciriello 1991). Jorgensen et al (1999) examined subjects performing sagit­ tal lifting activities and correlated the psychophysical limits with both calculated biomechanical and measured physio­ logical values. They observed that the decisions made by subjects when increasing and lowering weights towards a MAWL appeared to correlate with both the biomechanical and physiological parameters. They felt that the psy­ chophysical approach allowed lifters to address more of the risks associated with all parts of the body rather than those specific to the low back as seen in the traditional biome­ chanics approach.

There are very few tasks performed in the work, home or recreational environments that do not involve manual han­ dling of some description, whether involving relatively low weights (pens, television remote controls, etc.) or the larger loads handled in the heavy industries such as the mining and foundry industries. Manual handling is a term used to describe any activity that involves the generation of physi­ cal force by the person to complete the task - pushing, pulling, carrying, lifting, etc. This review will limit itself to the manual handling task of lifting. All lifting tasks share common characteristics. Lifting involves the movement of an object from one location to another location, generally traversing both vertical and hor­ izontal distances, and can be subdivided into three stages:

hand(s) in a position on the load to allow control of the load during lifting. The need to access the load is the key driver for the posture that the body adopts at the commencement of the lift. Confined or cramped workspaces, for example aircraft luggage holds, will also affect the posture assumed in this stage. 2. Movement. Pure lifting - i.e. pure vertical movement of a load during lifting - is rare, with most lifting involving a dimension of horizontal movement. The direction of the horizontal movement should also be considered, as movements of the load in directions away from the sagittal plane will involve twisting and/ or asymmetrical loading of the spine. Successful completion of this stage will depend on generation of sufficient force by the musculoskeletal system and results in the development of increased stress on the spine. 3. Placement. At the completion of the lift the lifter must control the load to a set destination. Factors affecting this stage of the lift include the speed of lifting, the location of the destination, the nature of the load and the precision required in placing the load.

Access. The initial stage involves the lifter getting the

Biomechanical approach
Ethical and methodological constraints limit the capacity to measure internal loads on the body during manual han­ dling activities by direct measurement methods (Langrana et al 1990). The biomechanical loads on the lumbar spine are one of the contributing factors to the occurrence of low back pain (Langrana et al 1990). The biomechanical approach involves 'the systematic application of engineer­ ing concepts to the functioning of the human body to pre­ dict the distribution of internal musculoskeletal forces resulting from the interaction with externally applied forces of the task' (Waters & Putz Anderson 1996). The human body is considered to be a system of mechanical links, each of a known physical size and form and these dimensions are used to construct biomechanical models, which reduce the complexity of the system to enhance understanding (Chaffin et al 1999). The complexity of the mathematical formulation and ease of use of the biomechanical models vary significantly between the different models. Important considerations when using or interpreting the findings from a biomechanical model are (Waters & Putz Anderson 1996):
• • •


There are three approaches traditionally used to study the effect of lifting on the human body:

Physiological approach
This approach examines the physiological demands (heart rate, 02 consumption, ventilatory rate, EMG and blood lactate levels) of lifting on the human body. The determinants of safe lifting in this approach include the minimization of the energy demands on the lifter, reduc­ ing the accumulation of physical fatigue that may con­ tribute to musculoskeletal injury (Waters & Putz Anderson 1996).

the mechanical nature of the model (static vs dynamic) dimensionality of the model (two- or three-dimensional) accuracy of the representation (single or multiple muscles, lAP (intra-abdominal pressure), muscle co-contraction, active and passive elements) complexity of the input needed to use the model (mechanical parameters, physiological measures of muscle function, musculoskeletal geometry). -

Clinical biomechanics of lifting


From an engineering mechanics perspective, in a three­ dimensional modelling system the complexity of the input data which can be accepted by the system will often be lim­ ited by its mathematical capacity (Langrana et al 1990). Early models used simple vector moments, incorporating simple lines of pull to represent the muscular elements in the model. Given the cross-sectional dimensions of the muscles and the dynamic nature of their recruitment this limited the accuracy of the models in predicting internal spinal loads (Davis & Mirka 2000). The use of more com­ plex modelling systems has improved the accuracy of the model outputs; however, this remains an area of concern when defining the validity of any modelling system. In general, for clinical purposes, a biomechanical model need only be as complex as is necessary to accurately and reasonably describe the nature of the loads occurring in the lumbar spine due to a particular work task, and often involves a trade-off between criteria of accuracy and realism versus simplicity and ease of use (Granata & Marras 1996). Decisions on safe lifting limits are made by comparing the internal stresses calculated using biomechanical mod­ els, with the experimentally induced failure loads of spe­ cific spinal tissue. If the computed internal stresses that result from the application of a known external load fall under the experimentally induced failure load of the spinal tissue, then the lift is considered to be 'safe'. When the cal­ culated internal stresses exceed the capacity of the tissue then it is hypothesized that injury will occur. Biomechanical models can then be used to develop or support risk control strategies that minimize the calculated stresses, allowing a safety zone during manual handling activities.


Biomechanical Physiological Psychophysical

� 1;] -g




Q) a:






Frequency (lifts/min)

Figure 8.1 Exam ple of conflicts amo n g bio m ech anical , psychophysical and physiological criteria. Reproduced with permission from Ayo u b Et Woldstad 1999.

makes it difficult for the clinical practitioner to make a deci­ sion on proper safety limits for manual handling, as demon­ strated in Figure 8.1. An attempt has been made to circumvent this problem with the development of inte­ grated models. These models involve a unique approach that considers all three of the primary stress measures - bio­ mechanical, physiological and psychophysical. A prime example of this approach is the revised National Institute for Occupational Safety and Health (NIOSH) lifting equation (Tables 8.1, 8.2, 8.3) (Waters et al 1994). The NIOSH lifting equation used population norms from the three approaches to develop the lifting model. The norms include:
Biomechanical: predicted maximum compressive forces on the L5/S1 should not exceed 3.4 kN. 2. Physiological: metabolic energy expenditure rates should not exceed safe limits (Table 8.4). 3. Psychophysical: safe limits should comply with the maximal acceptable weight limits of 75% of women and 99% of men.

Integrated approaches
It is not surprising, given the different approaches used, that calculated safe lifting limits may conflict between the approaches (Dempsey 1998, Ayoub & Woldstad 1999). This

Table 8.1

The revised NIOSH lifti n g equ ation (adapted from Waters et al 1 994)
Revised N IOSH lifting equation RWL














Key to revised N IOSH lifting equation
Lifting task descriptor

Source 23 kg, 226 N 25/H 1 - (0.003 [V - 75]) 0.82 + (4.5/0) 1 - (0.0032 A) See Table 8.2 See Table 8.3

Recommended weight l imit (RWL) Load constant (LC): the maximum value for RWL Horizontal multiplier (HM): rel ated to horizontal distance from hand grip to body Vertical multiplier (VM): related to height of load from ground level Distance multiplier (OM): related to distance load moves vertically Asymmetry multiplier (AM): related to the angle of asymmetry from the mid-sagittal plane Frequency multiplier (FM): related to the number of l ifts per minute Coupling multiplier (CM): rel ated to the quality of the persons coupling with the load



Table 8.2 Frequency multiplier table for revised NIOSH lifting equation (reproduced with permission from Chaffin et al 1 999)
Work duration Frequency Lifts/min (F)

<1 hour V<0.75 1 .00 0.97 0.94 0.91 0.88 0.84 0.8 0.75 0.70 0.60 0.52 0.45 0.41 0.37 0.00 0.00 0.00 0.00 V>0.75 1 .00 0.97 0.94 0.91 0.88 0.84 0.8 0.75 0.70 0.60 0.52 0.45 0.41 0.37 0.34 0.31 0.28 0.00

>1 but <2 hours V<0.75 0.95 0.92 0.88 0.84 0.79 0.72 0.60 0.50 0.42 0.35 0.30 0.26 0.00 0.00 0.00 0.00 0.00 0.00 V>0.75 0.95 0.92 0.88 0.84 0.79 0.72 0.60 0.50 0.42 0.35 0.30 0.26 0.23 0.21 0.00 0.00 0.00 0.00

>2 but <8 hours V<0.75 0.85 0.81 0.75 0.65 0.55 0.45 0.35 0.27 0.22 0. 1 8 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 V>0.75 0.85 0.81 0.75 0.65 0.55 0.45 0.35 0.27 0.22 0.1 8 0.1 5 0.1 3 0.00 0.00 0.00 0.00 0.00 0.00

<0.2 0.5 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 >1 5

values in metres


Biomechanical criteria for determining safe lifting
In developing criteria for safe levels of lifting the NIOSH lifting equation used three biomechanical criteria based on a review of the literature. These criteria were: The joint between L5 and S1 is the level of greatest lumbar stress during lifting. 2. Compressive force at this level is the critical stress vector. 3. The compressive force criterion that defines increased risk is 3.4 kN (Waters et a11994).

It has been proposed that at a compressive force of 3.4 kN on the lumbar spine, vertebral end-plate microfractures

begin to occur in less than 1% of the male worker popula­ tion and less than 25% of female workers during occasional lifting tasks, while compression levels above 6.4 kN are hazardous to the majority of workers (Chaffin et a11999). The mechanism for spinal injury under compressive load has been reported as occurring through failure of the end­ plates of the vertebral bodies and underlying trabeculae as the nucleus pulposus bulges upwards and downwards (Adams et al 2000). The magnitude of the compressive forces during a single lift is unlikely to cause end-plate fail­ ure, and it has been proposed that injury of this type is most likely to be cumulative (Fig. 8.2) (Adams & Dolan 1995). Lotz et al (1998) studied the effects of sustained compres­ sion on the intervertebral disc structure of mice and identi­ fied that compression resulted in disorganization of

Table 8.3 Coupling multiplier table for revised NIOSH lifting equation (reproduced with permission from Chaffin et al

1 999)

Table 8.4 Metabolic energy expenditure limits (Kcal/min) for lifting, as used in the revised NIOSH lifting equation (adapted from Rodgers et al 1 99 1 )
Duration of lifting (during a work day) Lift location (V) mm

Coupling multiplier Coupling type

< 0.75 m 1 .00 0.95 0.90

V > 0.75 m 1 .00 1 .00 0.90

< 750 4.7 3.7 3. 1

V > 750 3.3 2.7 2.2

Fai r

< 1 hour
Between 1 and 2 hours Between 2 and 8 hours


Clinical biomechanics of lifting


Biomechanical forces (internal and external) resulting in compressive loads> 3400 N

Vertebral end plate microfractures

Scar tissue formation over the area of microfracture

et al 1995, Ayoub & Woldstad 1999, Fathallah et al 1999). Shear forces in particular have been identified as poten­ tially contributing to the risk of low back injury. However, safe limits for shear force exposure have not been as well established as those for compressive loads (Karwowski et al 1991, Davis & Marras 1998). Davis & Marras (1998) pro­ posed a shear tolerance limit of 1 kN at which point there was an increased risk of tears of the annulus fibrosus; how­ ever, extensive work is still required in this area before shear tolerance values can be used as the biomechanical cri­ terion for determining safe lifting limits.

Style of lifting
Reduced nutrient delivery to the disc (due to the scar tissue layer)

Gradual annulus fibrosus degeneration

Lifting from below waist height is characterized by ankle dorsiflexion, knee, hip and lumbar flexion during the 'access' part of the lift, followed by ankle plantarflexion, knee, hip, and lumbar extension to perform the lift (Burgess-Limerick 2001). There are three main lifting tech­ niques described in the literature, which involve different relative movements between the joints of the trunk and lower limbs (Fig. 8.3). The technique description pertains to the posture adopted at the start of the lift (Burgess-Limerick et al 1995, Burgess-Limerick 2001).
Squat lifting

Decreased tolerance and work

Low back pain


Figure 8.2 Proposed seq uence of events for spin a l degeneration following application of compressive forces. Adapted with permis­ sion from Ma rras 2001.

annular architecture, increased biomechanical instability (i.e. increased neutral zone), altered type 2 collagen and diminished cellularity. They reported that sustained com­ pression resulted in cell death in the nucleus and inner annulus, possibly due to the mechanical stress or the adverse biochemical environment from the resultant water loss. Their observed annular morphological changes and increased biomechanical instability are consistent with those reported for degenerative human intervertebral discs. It was proposed that the effects of sustained or repeated compressive loading of the discs will hasten the disc degen­ erative process through cellular and biomechanical mecha­ nisms (Adams et al 2000), even though they may fall below 'safe' biomechanical compressive levels. Critical reviews of the criteria for the determination of safe lifting limits by Leamann (1994) and Dempsey (1998) identified that the use of compressive forces as the biome­ chanical 'safety' criterion may be flawed and both authors concluded that further research was needed. Other biome­ chanical criteria that may be used include the external hip moment, the anteroposterior (AP) shear force, lateral shear forces and the kinematic parameters of the torso (Marras

At the commencement of the lift the body starts with a pos­ ture of ankle dorsiflexion, full knee flexion and hip flexion with the trunk maintained close to upright. Squat lifting can be further divided on biomechanical grounds into lift­ ing with a small-sized load, which can be lifted between the knees, and lifting a larger load, which must be lifted in front of the knees in the squatting position (Chow 2001). Changes in load dimensions, and hence capacity to lift between the knees during squat lifting, will affect the distance of the load from the body, a powerful influence on the resultant stresses on the spine during lifting.

Stoop lifting

This describes the other extreme of lifting where the knees are minimally flexed, the ankles maintained in plantar­ grade and the trunk near maximal flexion. It is also termed the 'derrick' lift due to its similarity to the actions of the derrick crane (Oborne 1995). This lifting style is character­ ized by maximum lumbar flexion at commencement of the lift.

Semi-squat lifting

The posture involved lies between the stoop and squat lift with moderate trunk and knee flexion. Semi-squat lifting has been reported as the most common type of lift adopted when free dynamic lifting, with either of the two extremes of lifting styles rarely used when asked to perform free dynamic lifting, particularly over an extended period of time (Gagnon & Smyth 1992, Burgess-Limerick et al 1995, Burgess-Limerick & Abernethy 1997).




Fig u re 8.3

Three different lifting styles. A: Stoop lifting. B: Sq u at lifting. C: Semi-squat lifting.

A problem with defining the lifting styles by the posture demonstrated at the initiation of the lift is that it does not control for the movement pattern that the person uses when lifting (Burgess-Limerick et al 1 995, Hsaing & McGorry 1997). When a person uses different lifting strate­ gies there are changes in the coordination of the body and limb movements and in the motion pattern of the external load (Hsaing & McGorry 1997). During squat lifting, the lifter has a number of different strategies available to lift the load. They may pull the load closer to the body during the prelifting phase, use the body to jerk up the load during the lifting phase and then slide the load forward midway through the lift, or pull the load close to the body and develop a combined upward and forward momentum of the load before guiding it to touchdown (Hsaing & McGorry 1997) (Fig. 8.4). When teaching correct lifting we need to consider the motion patterns used in the lift as well as the initial posture. Hsaing & McGorry (1997) demon­ strated that manipulation of the motion patterns of the load could be used to 'control' the estimated compressive forces on the lumbosacral joint, with the latter combined style

motion pattern demonstrating the lowest increase in com­ pressive values. The pros and cons of each lifting style will depend on the biomechanical stresses that the lifting style places on the lifter 's trunk. In biomechanical terms, the main effect of dif­ ferent lifting styles will be on the magnitude and orientation of the moment of the load through affecting factors such as the object's centre of gravity, weight distribution (Obome 1995) and the posture of the spine during the lift. Trunk pos­ tures during lifting have been shown to be associated with the risk of low back pain (Granata & Wilson 2001).

Spinal motion segment
It has been reported that 85-95% of all disc herniations related to manual handling occur at the L4/5 and L5/S1 spinal levels and the L5/S1 level sustains the largest amount of force (Chaffin et al 1999). Tichauer (1971) pro­ posed that the load moment around the LS/S1 joint form the basis for setting safe limits for lifting and carrying. As described, the compressive force at this level has been used in setting biomechanical criteria. A range of other forces also act on the lumbar motion segment during lifting and these are shown in Figure 8.5. Lifting, with the concomitant development of flexor and extensor moments on the spine, results in development of both compression and shear forces over the motion segment (Burgess-Limerick 1999). In full trunk flexion, as occurs in stoop lifting, 70% of the resistance to further lumbar flexion is provided by the intervertebral ligaments (in particular the short ligaments of the apophyseal joints) while the disc resists only 30% of the flexion torque. Once we move past the elastic limits of the ligaments, the interspinous and supraspinous liga-

A Mobilization

B Stabilization

C Optimal strength utilization

Figure 8.4 Different lifting motion patterns. Reprod uced with permission from Hsiang & McGorry 1 997.

Clinical biomechanics of lifting




Anterior/posterior shear


Lateral shear

Forces acting on the spinal motion segment d u ri n g l ifting. Reproduced with perm ission from Marras 2001.

Figure 8.5

ments are damaged first (Adams et al 2000). Increased intradiscal pressure in this posture occurs due to tension in the posterior intervertebral ligaments and the posterior annulus. When we lift there is an increase in the compressive forces on the lumbar motion segments through an increase in the magnitude of the load moment acting on the spine and an increase in muscle activity used to raise the load. As the vertical spacing of adjacent vertebrae is small compared to their length and width, small changes in the angle of motion segment flexion can lead to large changes in the dis­ tribution of stress in the motion segment, with this effect being exaggerated with pathological changes and creep loading (Adams et a1 2000). When a cadaveric disc is loaded to reduce disc height by 20% (to simulate normal diurnal variation seen in vivo), the pressure in the nucleus falls by 36% while peaks of compressive stress rise in the annulus. Full lumbar flexion significantly increases the compressive pressure in the anterior annulus, while mid-flexion tends to equalize the compressive force across the whole disc (Adams et a1 2000). Young, well-hydrated discs are less sen­ sitive to changes in posture, and stress concentrations are only evident at the end of range (Dolan & Adams 2001). The apophyseal joints show similar changes secondary to nar­ rowed disc spaces with peak compressive forces in the apophyseal joints changing from middle to upper regions in the flexed posture to the inferior margins in lordotic pos­ tures (Dolan & Adams 2001). With the application of compressive force on the motion segment in a neutral position, the intervertebral disc pro­ vides the majority of the resistance. The facet joints provide little stiffness to compression in the neutral posture due to their vertical alignment; however, in the lordotic posture, such as in squat lifting, the facet joints can resist from 15 to 25% of the applied compressive load (Yang & King 1984),

which increases further in the presence of facet degenera­ tion and/or disc narrowing (Dunlop et al 1984, Yang & King 1984). Three factors can increase the amount of com­ pression force borne by the neural arch: pathological disc narrowing, prior long-term creep loading; and lordotic pos­ tures. When all factors are in place, the neural arch can resist up to 70% of the compressive stresses in the lordotic posture (Adams et al 2000). Biomechanically the properties of the intervertebral disc are influenced by its geometric parameters, such as height and area. The height and area of the disc vary between disc levels, between different people and within the same disc itself. There is a decrease in disc height from the fifth decade of life while the disc area increases with age (Natarajan & Andersson 1999). Within the same person, the disc varies during the day due to diurnal variations, with a loss of height, particularly in the first few hours of the day and related to severity of loading of the spine. This diurnal change of disc height results in changing of the load­ sharing capacity of the spinal elements during the day (Natarajan & Andersson 1999), with the disc taking more of the stress during flexion earlier in the day. Adams et al (1990) reported an increase in compression stiffness and more flexibility in flexion with diurnal changes. During the application of anterior shear forces to the motion segment, as occurs with trunk flexion, the apophyseal joints provide the majority of the resistance to further anterior shear through the development of com­ pressive stresses between the overlapping articular sur­ faces (Langrana et al 1990). In the general population there is wide variation between the anatomical orienta­ tion of the apophyseal joints of the lumbar spine (Bogduk 1997). In flexed lumbar postures, the apophyseal joints provide resistance to further flexion through passive stretching of the capsular fibres, but the capacity of the joints to resist anterior shear forces will depend on the orientation of the articular surfaces. Apophyseal joint articular surfaces parallel to the sagittal plane are less likely to be able effectively to resist significant increases in anterior shear forces that may develop from lifting in a flexed posture (Bogduk 1997). This may place greater anterior shear stress on the intervertebral disc, a plane that it is not well designed to resist, increasing the poten­ tial for injury to this structure. A factor not always considered in biomechanical model­ ling, but one that has significance clinically, is the effect of creep on the motion segment. Human biological tissue has a viscoelastic nature and when subject to static or repeated postures undergoes creep, with a reduction in stiffness of the passive tissues of the motion segment (Best et alI994). Viscoelastic creep has been demonstrated following cyclic and prolonged loading in flexion and has been shown to increase the laxity of the intervertebral joint, leading to high rates of instability, injury and low back pain in individuals involved in lifting (Gedalia et al 1999). Injuries associated



with spinal instability can reportedly occur at compressive forces approaching 88 N (Granata & Wilson 2001). It has been reported that the risk of low back injury is increased when lifting is performed many times during the day (Mundt et al 1993). The two mechanisms proposed include the altered muscle activation patterns found during fatiguing work activities, which may result in increased spinal compression, or the resultant muscle insufficiency, which may shift the loading to the passive tissue of the body. Sparto & Parnianpour (1998) used EMG-assisted bio­ mechanical modelling to demonstrate minimal increase in spinal compression as a result of the changing muscle recruitment patterns and suggested that the injury mecha­ nisms that result from repetitive or sustained posturing may be due to the change in the viscoelastic passive tissue responses or muscular insufficiency. This raises the potential for reduced capacity of the pas­ sive structures of the spine to resist extra stresses, poten­ tially resulting in temporary instability of the motion segment in that posture, increasing the risk of injury. The motion segment therefore relies more heavily on the dynamic components of motion segment stability, the abdominals and erector spinae to overcome the stress of any imposed loads (Solomonow et aI 1999). Deficiencies in this dynamic stabi­ lizing system may result in risk of injury below 'normal' biomechanical failure criteria. The deformation and reduced thickness of the disc are thought to increase the laxity of the joints, increasing the range of IV movement as well as instability and injury (Solomonow et aI 1999). Creep loading causes the annulus to resist a lower proportion of the bending moment applied to the spine and the ligaments to resist more. This implies that the annulus resists bending most strongly in the early morning when the disc is hydrated (Dolan & Adams 2001) and less during the day as the spine is subject to creep.

Intradiscal pressure
The pressures increase within the intervertebral disc during all manual handling activities (Kroemer & Grandjean 1997). Nachemson & Elfstrom (1970), in a review of intradiscal pressures during different lifting postures, identified that there was a sharp rise in intradiscal pressure at the level of L3 /4 during stoop lifting compared with squat lifting when lifting a 20 kg load. When a load is held at a distance from the body, as in lifting a load in a squat lift around the knees, there is a significant increase in compressive forces at the lower lumbar levels, further increasing intradiscal pressure (Kroemer & Grandjean 1997). This increase in intradiscal pressure results from the increased muscular activity and lumbar flexion used in the lift posture.

Muscle activity
The erector spinae act to produce the extensor moment required to overcome the weight of the load and extend the

trunk to the upright posture during lifting. For biomechan­ ical modelling purposes, the action of the erector spinae muscles can be represented as a force moment acting on the spinal motion segment. Generally, this force moment has been represented as acting with a moment arm of 50 mm, a value resulting from the early work of Bartelink et al (1957). This value has been questioned with more recent work indicating that the moment arm of the erector spinae will vary between different lumbar postures. Tveit et al (1994) reported that the moment arm of the erector spinae increased by approximately 15% when the lordosis was increased, increasing the mechanical efficiency of the erec­ tor spinae. It was also reported that the upper lumbar and lower thoracic erector spinae portions of the erector spinae may contribute to the resultant extensor moment through their action on the erector spinae aponeurosis (or superfi­ cial dorsal tendon). This mechanism could theoretically increase the moment arm of the erector spinae to a maxi­ mum value of 85 mm, although this will depend on the spe­ cific anthropometric characteristics of the lifter and the posture assumed. Extreme lumbar flexion postures are characterized by the absence of EMG activity in the lumbar erector spinae (McGill & Kippers 1994, McGorry et al 2001), termed the flexion-relaxation response (FRR). A similar reaction has also been demonstrated in the hamstring muscles (McGorry et al 2001). Lifting with a lordosis, such as in squat lifting, was shown to result in earlier peak EMG readings in the erector spinae than lifting with the lumbar spine in kypho­ sis. During stoop lifting, the FRR was evident and the peak EMG response was delayed towards the middle of the lift (Holmes et al 1992). While the torque values around the spinal motion segments were similar between the two lifts, the orientation of the motion segment and its capacity to resist the imposed forces were different. The lack of erector spinae activity which occurred early in the stoop lift (i.e. FRR) results in the flexed spinal motion segment resisting the flexor moment by the posteriorly placed passive struc­ tures, including the paravertebral ligaments, interspinous ligaments, posterior fibres of annulus fibrosus, and the pas­ sive elements of the muscular system. Segmental muscle recruitment in the erector spinae mus­ cles progresses in the caudad-cephalad direction during trunk extension from full flexion, independent of the speed of lifting (McGorry et al 2001). Solomonow et al (i998) iden­ tified a primary reflex arc between the mechanoreceptors in the spinal ligaments and facet joint capsules to the multi­ fidus muscle. This reflex arc was triggered following the application of tensile loads to the spinal ligaments and resulted in contraction of the multifidus muscle at the level of ligament deformation and at one level above and /or below. This activity reached a peak when the stress in the ligament approached moderate levels that could 'poten­ tially cause damage to the ligament tissue (Solomonow et al 1998). This reflex arc appears to be present to protect the passive tissue constraints of the spine towards the end

Clinical biomechanics of lifting


range of lumbar flexion, although the presence of the FRR would suggest that this reflex arc is overridden at the extremes of range. Experimentally based research, primarily on feline spines, has shown that this reflexive muscular activity decreased during cyclic activity because of desensitization of the mechanoreceptors in the viscoelastic structures as they become subject to laxity (Solomonow et aI 1999). This was observed to occur even before fatigue of the erector spinae muscles set in. Gedalia et al (1999) observed that after 50 minutes of cyclic loading on the feline spine, recov­ ery of this reflex arc did not appear to occur after 2 hours of rest. Taimela et al (1999) identified that there was a decrease in the capacity of human subjects to sense a change in lumbar position (proprioception) following lumbar fatigue activities in both control and low back pain patients, although this was significantly worse in LBP patients. The desensitization of the mechanoreceptors in the passive spinal tissues following repeated loading, as seen in feline spines, is an attractive mechanism to help explain the increased risk of low back pain following repeated manual handling. It could be hypothesized that following the cyclic loading of the passive intervertebral tissue resulting from repeated manual handling the human spine is more vulnerable to injury due to reduced neuromuscular control. This remains an exciting area for further research. Contraction of the erector spinae muscles (in particular the pars lumborum fibres of the longissimus thoracis and iliocostalis lumborum) results in the development of a pos­ terior shear force on the superior vertebrae. This has the potential effect of reducing the effect of anterior shear forces generated by the weight of the upper trunk and load (Burgess-Limerick 2001), but this capacity to resist anterior shear forces will depend on the lumbar posture used. The erector spinae (longissimus thoracis and iliocostalis lumbo­ rum) in the flexed posture have changed lines of action rel­ ative to the motion segment (by changing the cosine of the orientation of the line of action) and are therefore less able to resist the anterior shear forces seen to cause damage to the spine in full flexion (McGill et al 2000). Other muscles (multifidus, quadratus lumborum, psoas) also resist ante­ rior shear and would appear to be less affected by the angle of trunk. Despite the well-developed extensor muscles of the lum­ bar spine, biomechanical modelling indicated that the cal­ culated extensor moments to be overcome at the lumbar spine when lifting heavy loads exceeded the demonstrable capacity of the erector spinae (Gedalia et aI 1999). This sug­ gested that other mechanisms must assist the activity of the erector spinae muscles in generating sufficient extensor moment to overcome the applied load. Gedalia et al (1999) provided an excellent review of the various perspectives put forward to explain the discrepancy between calculated and actual forces generated. Theories include the arch the­ ory, where the lumbar spine is viewed as an arch braced by the intra-abdominal pressure (lAP), the hydraulic amplifier

theory, where the thoracolumbar fasciae surrounding the muscles act to brace the erector spinae muscles, increasing their power, or the passive posterior musculoligamentous system. In this latter system the passive ligamentous sys­ tem and the passive tension generated in the erector spinae muscles was used to overcome the load early in the lift, until the moment arm of the load was sufficiently reduced as the trunk approached the erect posture for the active ten­ sion of the erector spinae muscles to take over. Marras et al (2001) identified that patients with low back pain had higher resultant spinal compressive loads during free dynamic lifting despite reducing their effective trunk flexion moments by restricting their flexion range of motion and speed of movement. This increased spinal compressive load resulted from the increased levels of muscle coactiva­ tion demonstrated in this group, particularly when lifting below waist height. Another important factor was the influ­ ence of body weight, which Marras et al (2001) reported had a significant effect on increasing the spinal compressive load.

Intra-abdominal pressure
The concept that pressures within the trunk may assist with the mechanical efficiency of the trunk during lifting was first proposed in the 1920s. The original hypothesis was that the flexion moment created by the application of a load anterior to the axis of rotation of the motion segments would be counteracted by development of pressure in the trunk cavities (Chaffin et aI 1999). It was hypothesized that this would reduce the activity required of the erector spinae muscles, reducing the stress on the vertebral column. Early work by Bartelink (1957) and Morris et al (1961) concluded that there would be a 30% reduction in stresses over the lumbosacral joint with the development of intra-abdominal pressure (lAP). Recently this hypothesis has been brought into question as a result of extensive laboratory based work in this area. Intra-abdominal pressure responses appear to be divided into an initial peak response at the commence­ ment of the lift, a lower sustained pressure while the load was being raised and a further peak associated with the placement of the load. Interestingly, Hamberg et al (1978) used systematic strengthening exercises for the abdominal muscles and reported that while there were measurable increases in strength of the abdominal and back muscles these did not equate into increases in lAP while the subjects were lifting loads. How the lAP was generated may also affect the biome­ chanical influence on the spine. When developed as a result of the Valsalva manoeuvre, the increase in lAP was accom­ panied with an increase in back extensor muscle activity which resulted in increases in spinal compression forces, as measured by disc pressure measurements and from biome­ chanical modelling (McGill & Norman 1987). The role of lAP in lifting requires further clarification (Chaffin et al 1999). McGill & Norman (1987) and Marras et al



(2001) concluded that the co-contraction of several abdom­ inal muscles (in particular the transversus abdominis and the oblique abdominals) acts to stiffen the torso, reducing the neutral zone, but also increasing IAP. It has been sug­ gested that the muscle tensions involved would cancel any major unloading of the spinal disc due to increases in IAP (McGill & Norman 1987), hence the IAP has been depicted as a by-product of antagonistic co-contraction of the torso muscles to stabilize the spine during the act of lifting (Cholewicki et al 1999). Hodges et al (2001) have raised some questions about this proposal, suggesting that increases in IAP may in fact facilitate an extensor torque if the IAP is generated through selective muscle recruitment, in particular of the diaphragm, pelvic floor muscles and transversus abdominis (Hodges et al 2001). advocating the squat technique to prevent low back pain. They reported that the positive effects for squat lifting with respect to estimated spinal force moments and compression values were found only when the squat lift allowed lifting from a position between the feet, reducing the load on the low back by up to 30%. Issues with squat lifting include the higher ground reaction forces due to the greater vertical excursion of the body centre of mass, which are often ignored in static biomechanical modelling. They reported that in lifting tasks where the load was not lifted from a position between the feet, the net moment and compressive load through the lumbar spine were lower in stoop lifting. In contrast the shear and bending moments were higher in stoop lifting. Straker & Duncan (2000) found that subjects reported more discomfort and lower MAWL during squat lifting a medium-sized box from floor to knuckle height than with the stoop lift. It appears therefore that there is no clear-cut advantage offered by one extreme lifting style over the other. This is reflected in the clinical observation that subjects choose the semi-squat lifting style during free dynamic lifting rather than squat or stoop lifting.

Other joints
Biomechanical modelling of dynamic lifting has shown that the forces over the hip joint can be quite large, particularly when the load cannot be held close to the body. The capac­ ity of the hip muscles to generate sufficient force to over­ come the flexor moment generated by lifting loads is well documented (Farfan 1978, Bogduk 1997). Unfortunately the strong hip extensor muscles are only able to rotate the hip and pelvis backward on the femurs, leading to increased flexor moments acting on the lumbar spine (Bogduk 1997). The strong one-joint hip flexor muscles are less likely to be affected by lifting posture than the longer multijoint mus­ cles. During lifting from the semi-squat posture the interac­ tion between knee and hip extension allows the hamstrings and quadriceps to work together to maintain an adequate length-tension relationship facilitating their effectiveness, a situation that is less likely to occur during stoop lifting (Burgess-Limerick 1999). In the squat lift position, the knees are in a 'close-packed' position, and the heels are generally off the ground. This places the body in an unstable position and places greater stress through the knees during the early part of the lift. Perturbations of the load during the lift may be less able to be withstood due to the relative instability of the body, increasing the potential for asymmetrical stresses through the lumbar spine. Postures of full knee flexion are generally discouraged in patients to avoid the significant patellofemoral joint compression that results from this pos­ ture, further exacerbated when a load is lifted. The patellofemoral joint is an area commonly involved in osteoarthritic changes in the ageing population. Stoop lift­ ing and semi-squat lifting place less stress through the knee joints, allowing these joints to avoid the close-packed posi­ tions. van Dieen et al (1999) presented an excellent review of the biomechanical evidence in support of advocating the squat lift compared to the stoop lift as a control measure to prevent low back pain. They concluded that the biome­ chanical literature did not provide substantial support for

In considering the biomechanical effects on the spine of the different lifting styles, we need to consider the range of other factors that may influence the effect on the spine. These factors can be divided into three main categories and are listed in Table 8.5.

Personal risk factors
These are the characteristics of the worker that may affect the probability that an injury may occur. Both age and gender have been shown to affect the bio­ mechanical characteristics of the spine Gager & Luttmann 1992). Age will affect the mechanical behaviour of the spinal motion segments, secondary to degenerative changes, as well as reducing the strength of the trunk mus­ cle forces available to resist the internal pressures when lift­ ing (Stubbs 1985). Gender differences are based on differences in anthropometric characteristics between male and female population groups which will affect trunk weight, centre of mass and muscle moment arms. It has also been suggested that differences in lumbar lordosis angles between genders will affect spinal stability during lifting (Granata & Orishimo 2001). In the clinical application of risk management strategies to address risks associated with lifting, it is often difficult to address the personal risk factors. The basic tenet of ergonomics to 'fit the task to the person' is the safest guide when undertaking risk management programmes. Behavioural health programmes aiming to improve muscu­ loskeletal and cardiovascular health and fitness, facilitate smoking cessation and improve workplace morale may be useful in reducing the risks associated with lifting'activities.

C l i n i cal b i omecha n i cs of l ifting


Table 8.5 Risk factors associated with manual handling (adapted from Stubbs 1 985 and Waters 8: Putz Anderson 1 996)
Personal f acto rs
E n v i r o n men t al f acto rs

Job rel ated f acto rs Task Load

• •

• • • • • • • • • •

Sex Anthropometry (body weight and height) Physical fitness and training Lumbar mobility Strength Medical history Years of employment Smoking Psychosocial factors Anatomical abnormalities Ski l l levels Clothing worn

• • • • • •

Humidity Light Noise Vibration Foot traction Space available

• •

• •

location of l oad relative to worker. Reach and height Distance object is to be moved Frequency and duration of handling activity Bending and twisting Postu ral requirements, preceding and during l ift

• • • • • •

Weight of object or force required to move the object Stability of load Depth of l oad Centre of g ravity Breadth Height of l oad Height of l oad

However, they should only form a part of a total risk man­ agement strategy.

Environmental risk factors
These are conditions or characteristics of the external sur­ roundings that may affect the probability of an injury. Issues such as the quality of the floor surface upon which the lift is to be performed, the ambient environment and the space available in which to perform the lift will all affect the risks associated with lifting activities.

and trunk sagittal angle. Other authors have identified fac­ tors such as asymmetry, speed of lift and horizontal and vertical position of load and load mass (van Dieen et al 1999). As described, the NIOSH lifting equation has identified a number of different physical parameters that need to be considered when analysing a lift. The effects of job related risk factors are briefly described.

Horizontal position of the load
The horizontal position of the load relates to the position of the centre of mass of the load relative to the axis of rotation of the motion segment in the horizontal plane. The NIOSH lifting formula has defined the minimal distance that the centre of mass can be held from axis of rotation of the spine as 250 mm, which takes into account the abdominal cavity. Changes in the horizontal position of the load will have a dramatic effect on the moment of the load, significantly affecting spinal compression values. The increase in moment magnitude is non-linearly related to the increases in horizontal position of the load with an increasing rate of increase in moment magnitude as the load moves further away from the body (Schipplein et al 1995). As the load moves away from the body, the lever arm of the load acting at the spinal level increases, magnifying the flexor torque produced at the spinal level. The spinal exten­ sor muscles, working at a relatively fixed lever arm, must work significantly harder to balance the load. The increased activity of the extensor muscles result in increased compres­ sive loads over the underlying motion segments. Chaffin et al (1999) have recommended that the minimization of the horizontal distance of the load is the most important control mechanism when considering the biomechanical effect of lifting on the body. Figure 8.6 describes the predicted 15/S1

Job related risk factors
These are the characteristics of the task that may affect the likelihood of an injury and are usually considered the most important in biomechanics as they directly affect the mag­ nitude of the physical hazard to the worker (Waters & Putz Anderson 1996). They are also the easiest to measure and change in the occupational arena. However, consideration of just one of these factors - i.e. load mass - may underesti­ mate the effect of the lift on the lumbar spine (Davis & Marras 2000). Changes in load weight may lead to changes in trunk dynamics, which may offset any of the benefits of the reduced load weights. It is therefore more important to consider how the person interacts with the load rather than the actual weight of the load itself. Marras et al (1993, 1995) studied the contribution of var­ ious biomechanical workplace factors to the risk of low back injury in over 400 manual handling jobs in 48 different industries. They identified that the combination of five trunk motion and workplace factors were best associated with the risk of low back injury using multiple logistic regression modelling. These included lifting frequency, load moment, trunk lateral velocity, trunk twisting velocity

1 00


Figu re 8.6 Predicted LS/S l compression forces for varying loads and d ifferent postures. Reprod uced with permission from Chaffin et a 1 1 999.

Load·to-LS/S 1
5.0 4.0


20 cm

30 cm

40 cm

SO cm
500 N 400


N load N load N load N load

Predicted compression force (KN) on LS/Sl disc

Niosh 3400 N disc - - -- c m r s�o l mTt -- - -- -- - o pe n i

300 200 100

3.0 2.0 1.0

No load






Load horizontal H distance from LS/S 1 disc (average male anthropometry in postures above)

compressive forces for specific loads under different hori­ zontal distances from the spine. Increases in the horizontal distance of the load will not only increase the spinal compressive forces but it will also reduce the strength capacity of the subject (Kumar & Garand 1992), increasing the potential for injury in these postures (Kumar 1996). Furthermore, in a study of the effect of changes in horizontal distances of the load during peak exertions in stoop and squat lifting, Kumar (1996) found that reaching between full, three-quarters and half horizon­ tal reach distances had significant effects on the strength capacity of the lifter.

Lifting frequency
Increasing the frequency of the lift has been shown to have effects on safe lifting levels in both the physiological and psychophysical approaches. Mirka & Kelaher (1995) stud­ ied the effects of different lifting frequencies (between three and nine lifts per minute) on the kinematics of the trunk when free dynamic lifting. They reported that the higher frequencies of lifting resulted in higher levels of sagittal trunk acceleration, particularly between three and six lifts per minute. This occurred despite the fact that the frequen­ cies used did not result in a state of continuous lifting, i.e. even at nine lifts per minute the subject had time between lifts to rest (Fig. 8.7). This was supported by Nussbaum et al (1997) who reported significant increases in spinal com­ pression values, using an EMG-assisted biomechanical model, when lifting rates were increased 20% from pre­ ferred 'comfort' rates. Increases in lifting frequencies are biomechanically prob­ lematic for the spine when they increase the speed of the lift. This has been shown to increase the load moment act­ ing on the spine (Lavender et aI 1999), increasing the spinal compression values (Mirka & Kelaher 1995) and placing the spine at greater risk of injury (Marras et aI 1995). An interesting observation from Mirka & Kelaher's study was that, as the lifts continued over the 20 minute time span, the lifters demonstrated significant increases in trunk sagittal acceleration, although the time at which this occurred varied between subjects (Mirka & Kelaher 1995). The timing of this change in trunk acceleration corre-

Vertical position of the load
The height of the load relative to the lifter is a major driver behind the posture assumed when lifting, and hence the stresses through the body. Higher placed loads, such as with handles or on a raised stand, will reduce the degree of general flexion required to access the load. The less the degree of flexion required to access the load, the more likely the subject is to assume a neutral spine pos­ ture during lifting, reducing the biomechanical stresses through the spine, and facilitating trunk muscle activity (Tveit et a1 1994, McGill et aI 2000). The higher the load is placed vertically at the commencement of the lift, the shorter the vertical distance to be traversed during lifting, reducing the body's centre of mass vertical excursion, fur­ ther reducing the biomechanical stresses on the spine (van Dieen et aI 1999).

Clinical biomechanics of lifting

1 01




<= o

OJ '" (J)






200 -=F-------.---, 3 6

Lifting frequency (lifts/minute)

� 60 em ---0- 30 em Figure 8.7 Saggital tru n k acceleration va l ues with different levels of l ifting freq uency. Reprod u ced with permission from M i rka Et Kelaher 1 995.

sponded to change of subject lifting styles from squat type lifting to a stoop lift (Mirka & Kelaher 1995). This change in lifting style corresponds to the findings from the physio­ logical approach, which identified that the stoop lift was the most energy-efficient form of lifting.

Use of handles
The psychophysical approach has demonstrated that MAWLs were significantly higher for loads with handles compared to those without handles (Morrissey & Liou 1988, Drury et al 1989). Davis et al (1998) used an EMG­ assisted biomechanical model to calculate the internal forces on the spine when lifting crates with and without handles from a conveyor to a pallet. They found that the presence of handles reduced the anteroposterior shear and compression forces on the lumbar spine for loading all lev­ els of the pallet. They reported that this was due to the com­ bination of several factors including kinematic variables, muscle co-activity, and increased vertical heights resulting from the use of the handles. Lifting without handles resulted in greater activity in the antagonistic trunk mus­ cles, i.e. rectus abdominis, and greater maximum sagittal flexion of the trunk. This reflected the earlier findings of Kromodihardjo & Mital (1987) which reported that lifting loads with handles resulted in less biomechanical stress through the lumbar spine of the lifter.

the time delay between the onset of knee extension and lumbar extension, implying that the passive constraints to further flexion in the lumbar spine were used to overcome the load moment for longer with heavier weights, before active lumbar extension occurred. Davis & Marras (1998) found that several kinematic vari­ ables were significantly affected by changes in load weight when lifting loads over 20 kg but that overall postures were not. They reported that small changes in load weights dur­ ing free dynamic lifting had a limited effect on spinal loads due to changes in trunk kinematics. Subjects tended to lift with an increased velocity and tended to hold the load fur­ ther away from their body when lifting lower weights resulting in increased spinal loading. Davis & Marras (1998, 2000) expressed the ratio of computed dynamic compres­ sion values with the static compression values, calculated using static and dynamic biomechanical models (Fig. 8.8). They found that as the weight increased the ratio of dynamic to static compression values dropped, indicating that the dynamic factors had a greater influence over the lower loads. Unfortunately this study limited itself to loads over 9 kg and further work is required to investigate the effect of lower loads on trunk kinematics. This study reinforced the need to consider the range of factors, not just the weight but also dynamic factors such as trunk kinematics and muscle recruitment patterns, when assessing the risks involved in lifting activities.

Asymmetry of the lift
Asymmetry during lifting has been identified as a risk factor for the development of low back pain (Marras et al 1995, Lavender et al 1999) and complex combined dynamic motions of the trunk commonly occur during many of the industrial activities that predispose workers to low back pain (Straker et al 1997). Asymmetrical lifting generates complex patterns of spinal loading, placing increased stresses on the lumbar spine, increasing intradiscal pressure and the stress on cap­ sular ligaments (van Dieen & Kingma 1999, Chow 2001). Other proposed mechanisms for this increased risk from

2.8 .

'" '"



2.6 2.4 2.2 2.0 1 .8 1 .6 1 .4 1 .2 1 .0 9.1

!!! E

Cl. 0 u

Weight of the load
The weight of the load will have obvious effects on the load moment acting on the lumbar spine during lifting. Increases in load mass generally result in increased muscle activity, spinal loading and ultimately an increased risk of injury. Scholz (1993) reported that the load mass signifi­ cantly affected the delay after the start of the lift before lum­ bar extension occurred. Increases in load mass increased



$2 .2 E 0

'" <= >-


1 1 .8
















41 .8


Box weight


Figu re 8.8 Ratio of dyna mic with static com pressive values d u r­ ing l ifting. Reproduced with permission from Davis Et Ma rras 2000.

1 02


asymmetrical lifting include the increased muscle force used during asymmetrical lifting either to generate sufficient torque to perform the lift or secondary to the lifter 's percep­ tion of increased difficulty with asymmetrical lifting result­ ing in increased antagonistic muscle activation (van Dieen & Kingma 1999). This increase in muscle activation will result in higher spinal compression values. Investigating only one component of the net resultant spinal forces (i.e. compression) and ignoring the other com­ ponents, as occurs in traditional biomechanical modelling, may lead to erroneous judgements regarding the biome­ chanical risk involved (Straker et al 1997). This may result in underestimation of the spinal loading factors by up to 40% (Fathallah et al 1999). The combination of lateral shear and compression loading patterns appear to be the key to distinguishing the risk involved, particularly during asym­ metrical lifting tasks. There is a need to further develop tol­ erance of spinal tissue to dynamic complex loading conditions (Fathallah et al 1999). Asymmetrical lifting has also been shown to reduce the lifter 's trunk extensor capability (Chow 2001) and increase the levels of antagonistic co-contraction of trunk muscles (Marras & Mirka 1992). This increase in muscle co-activa­ tion occurred in muscles other than the erector spinae or latissimus dorsi, which appeared to be consistently active in all asymmetrical movements. The increase in co-activa­ tion also occurred with increases in trunk torque and trunk velocity (Marras & Mirka 1992), resulting in increased spinal compression values. Sudden and unexpected loading during lifting has been shown to increase the spinal compressive loading via mus­ cular activation by up to 70% (Mannion et al 2000). It has been reported that the muscle activity demonstrated during sudden and unexpected lifting exceeded that which was required to overcome the load, dramatically increasing the compressive stresses on the spine. The increased weight of the load, due to inertial properties was shown to increase

the compressive effect on the spine minimally compared to the effect of this increased muscle activity (Mannion et al 2000). Interestingly when subjects lifted an underestimated weight (up to 10 kg) at a self-selected, low velocity this increased muscle activity was not seen, and maximum com­ pression values and lumbar angles were no different than when expected lifting of the same mass occurred (van der Burg & van Dieen 2001). This difference may be secondary to the timing of the perturbation, with the low rate of appli­ cation of the perturbation during lifting an unexpected load allowing compensation to occur. Sudden increases in load mass during the lift may not allow compensatory mecha­ nisms requiring sudden increases in muscle activity to occur with the resultant increase in compression values.

Lifting remains a significant risk factor for the develop­ ment of low back pain in the community. When consider­ ing risk management strategies for the minimization of risks associated with lifting it is important to consider the range of parameters that affect the stresses on the lifter. It is no longer appropriate to consider manual handling training as a sufficient risk management strategy without assessing the whole person-task-environment system. The biomechanical approach, while not able to provide all the answers we need in the clinical environment, allows us to develop a better understanding of the influence of these parameters, and ultimately how we can change them to make the task of lifting safer.


biomechanics lifting

lumbar spine

Adams M A, Dolan P 1995 Recent advances in lumbar spinal mechanics and their clinical significance. Clinical Biomechanics 10: 3-19 Adams M A, Dolan P, Hutton W C et al 1990 Diurnal changes in spinal mechanics and their clinical significance. Journal of Bone and Joint Surgery 72: 266-270 Adams M A, Freeman B J C, Morrison H P et al 2000a Mechanical initiation of intervertebral disc degeneration. Spine 25(13): 1625-1636 Adams M A, May S, Freeman B J C et al 2000b The effects of backward bending on lumbar intervertebral discs: relevance to physical therapy treatments for low back pain. Spine 25(4): 431-437 Ayoub M M, Woldstad J C 1999 Models in manual materials handling. In: Kumar S (ed) Biomechanics in ergonomics. Taylor and Francis, Philadelphia Bartelink D L 1957 The role of abdominal pressure in relieving the pressure on the intervertebral discs. The Journal of Bone and Joint Surgery 39B(4): 71 8-725 Best T M, McElhaney J, Garrett W E et al 1994 Characterisation of the passive responses of live skeletal muscle tissue using the quasi­ linear theory of visco-elasticity. Journal of Biomechanics 27: 413-419 Bogduk N 1997 Clinical anatomy of the lumbar spine and sacrum, 3rd edn. Churchill Livingstone, Sydney Burgess-Limerick R 1999 Squat, stoop, or something in between? In: Straker L, Pollock C (eds) Proceedings of Cyberg 1999. Second International Cyberspace Conference on Ergonomics. International Ergonomics Association Press, Curtin University of Technology, Perth, Australia Burgess-Limerick R 2001 Lifting techniques. In: Kumar W (ed) International encyclopaedia of ergonomics and human factorS. Taylor and Francis, London Burgess-Limerick R, Abernethy B, Neal R J et al 1995 Self selected manual lifting technique: functional consequences of interjoint coordination. Human Factors 37: 395-411

Clinical biomechanics of lifting

1 03

Burgess-Limerick R, Abernethy B 1997 Toward a quantitative definition of manual lifting postures. Human Factors 39(1): 141-148 Chaffin D B, Park K S 1973 A longitudinal study of low back pain as associated with occupational weight lifting factors. American Industrial Hygiene Journal 34: 513-525 Chaffin D B, Andersson G B J, Martin B J 1999 Occupational biomechanics, 3rd edn. John Wiley and Sons, New York Cholewicki J, Juluru K, McGill S 1999 Intra-abdominal pressure mechanism for stabilising the lumbar spine. Journal of Biomechanics 32: 13-17 Chow D H K 2001 Lifting strategies. In: Karwowski W (ed) International encyclopaedia of ergonomics and human factors. Taylor and Francis, London Davis K G, Marras W S 1998 Is changing the box weight an effective ergonomic control? In: Proceedings of the Human Factors and Ergonomics Society, 42nd Annual Conference, Chicago, IL, Oct 5-9, Human Factors and Ergonomics Society Davis K G, Marras W S 2000 Assessment of the relationship between box weight and trunk kinematics: does a reduction in box weight necessarily correspond to a decrease in spinal loading. Human Factors 42(2): 195-208 Davis J R, Mirka G A 2000 Transverse contour modelling of trunk muscle distributed forces and spinal loads during lifting and twisting. Spine 25(2): 180-189 Davis K G, Marras W S, Waters T R 1998 Reduction of spinal loading through l�e use of handles. Ergonomics 41(8): 1155-1168 Dempsey P G 1998 A critical review of biomechanical, epidemiological, physiological and psychophysical criteria for designing manual materials handling tasks. Ergonomics 41 (1): 73-88 Dolan P, Adams M A 2001 Recent advances in lumbar spinal mechanics and their significance for modelling. Clinical Biomechanics 16(Suppl. 1): S8-S16 Drury C G, Deeb J M, Hartman B et al 1989 Symmetric and asymmetric manual material handling. 1: Physiology and psychophysics. ErgonOmics 32: 467-489 Dunlop R B, Adams M A, Hutton W C 1984 Disc space narrowing and the lumbar facet joints. Journal of Bone and Joint Surgery 66B: 706-710 Farfan H F 1978 The biomechanical advantage of lordosis and hip extension for upright activity: man as compared with other anthropoids. Spine 3: 336-342 Fathallah F A, Marras W S, Parnianpour M 1998 An assessment of complex spinal loading during dynamic lifting tasks. Spine 23(6): 706-716 Fathallah F A, Marras W S, Parnianpour M 1999 Regression models for predicting peak and continuous three dimensional spinal loads during symmetric and asymmetric lifting tasks. Human Factors 41(3): 373-388 Gagnon M, Smyth G 1992 Biomechanical exploration on dynamic modes of lifting. Ergonomics 35: 329-345 Gedalia U, Solomonow M, Zhou B et al 1999 Biomechanics of increased exposure to lumbar injury caused by cyclic loading. 2: Recovery of reflexive muscular stability with rest. Spine 24(23) : 2461-2467 Granata K P, Marras W S 1996 Biomechanical models in ergonomics. In: Bhattacharya A, McGlothin J D (eds) Occupational ergonomics. Marcel Dekker, New York, pp 115-136 Granata K P, Orishlmo K F 2001 Response of trunk muscle coactivation to changes in spinal stability. Journal of Biomechanics 34: 1117-1123 Granata K P, Sanford A H 2000 Lumbar-pelvic coordination is influenced by lifting task parameters. Spine 25(11): 1413-1418 Granata K P, Wilson S E 2001 Trunk posture and spinal stability. Clinical Biomechanics 16: 650-659 Hamberg J, Hemborg B, Holmstrom E, LOwing H, Moritz U, Nilsson M K, Akesson I 1978 Abdominal muscular activity and intra-abdominal pressure at different lifting techniques, before and after training of the abdominal muscles In: Proceedings of the XIX International

Congress on Occupational Health, Dubrovnik, Yugoslavia, Sept 25-30. International Commission on Occupational Health. Hansson T, Keller T, Spengler D M 1987 Mechanical behaviour of the human lumbar spine II: fatigue strength during dynamic compressive loading. Journal of Orthopedic Research 5: 479-487 Hildebrandt V J 1987 A review of epidemiological research on risk factors of low back pain. In: Buckle P W (ed). Musculoskeletal disorders at work. Taylor and Francis, London Hodges P W, Cresswell A G, Daggfeldt K et al 2001 In vivo measurement of the effect of intra-abdominal pressure on the human spine. Journal of Biomechanics 34: 347-353 Holmes J A, Damaser M S, Lehman S L 1992 Erector spinae activation and movement dynamics about the lumbar spine in lordotic and kyphotic squat lifting. Spine 17(3): 327-334 Hsaing S M, McGorry R W 1997 Three different lifting strategies for controlling the motion patterns of the external load. Ergonomics 40(9): 928-939 Jager M, Luttmann A 1992 The load on the lumbar spine during asymmetrical bi-manual materials handling. Ergonomics 35: 783-805 Jorgensen M J, Davis K G, Kirking B C et al 1999 Significance of biomechanical and physiological variables during the determination of maximum acceptable weight of lift. Ergonomics 42(9): 1216-1232 Karwowski W, Hancock P, Zurada J M et al 1991 Risk of low back overexertion injury due to manual load lifting in view of the catastrophe theory. In: Queinnec Y, Daniellou F (eds) Designing for everyone. Proceedings of the 11th Congress of the International Ergonomics Association, Taylor and Francis, London Kroemer K H E, Grandjean E 1997 Fitting the task to the human: a textbook of occupational ergonomics, 5th edn. Taylor and Francis, London Kromodihardjo S, Mital A 1987 Biomechanical analysis of manual lifting tasks. Journal of Biomedical Engineering 109: 132-137 Kumar S 1996 Spinal compression at peak isometric and isokinetic exertions in simulated lifting in symmetric and asymmetric planes. Clinical Biomechanics 11(5): 281-289 Kumar S, Garand D 1992 Static and dynamiC lifting strength at different reach distances in symmetrical and asymmetrical planes. Ergonomics 35(7/8): 861-880 Langrana N A, Edwards W T, Sharma M 1990 Biomechanical analysis of loads on the lumbar spine. In: Weisel S W, Weinstein J N, Herkowitz H et al (eds) The lumbar spine: the International Society for the Study of the Lumbar Spine, 2nd edn. W B Saunders, Philadelphia Lavender S A, Li Y C, Andersson G B J et al 1999 The effects of lifting speed on the peak external forward bending, lateral bending and twisting spine moments. Ergonomics 42(1): 111-125 Leamann T B 1994 Research to reality: a critical review of the validity of various criteria for the prevention of occupationally induced low back pain. Ergonomics 37(12): 1959-1974 Lotz J C, Colliou 0 K, Chin J R et al 1998 Compression induced degeneration of the intervertebral disc: an invivo mouse model and finite element study. Spine 23(23): 2493-2506 McGill S M, Kippers V 1 994 Transfer of loads between lumbar tissues during flexion relaxation phenomen. Spine 19(19): 2190 McGill S M, Norman R W 1987 Effects of an anatomically detailed erector spinae model on L4/LS disc compression and shear. Journal of Biomechanics 20(6): 591-600 McGill S, Hughson R L, Parks K 2000 Changes in lumbar lordosis modify the role of the extensor muscles. Clinical Biomechanics 15: 777-780 McGorry R W, Hsiang S M, Fathallah F A et al 2001 Tuning of activation of the erector spinae and hamstrings during a trunk flexion and extension task. Spine 26(4): 418-425 Magora A 1973 Investigation of the relation between low back pain and occupation. IV: Physical requirements: bending, rotation, reaching

1 04


and sudden maximal effort. Scandinavian Journal of Rehabilitative Medicine 5: 186-190 Mannion A F, Adams M A, Dolan P 2000 Sudden and unexpected loading generates high forces on the lumbar spine. Spine 25(7): 842--852 Marras W S 2001 Loads on the lumbar spine during dynamic work. In: Karwowski W (ed) International encyclopaedia of ergonomics and human factors. Taylor and Francis, London Marras W S, Mirka G S 1992 A comprehensive evaluation of trunk response to asymmetric trunk motion. Spine 17(3): 318-326 Marras W S, Lavender S A, Leurgans S E et al 1993 The role of dynamic three dimensional trunk motion in occupationally related low back disorders: the effects of workplace factors, trunk position and trunk motion characteristics on the risk of injury. Spine 18: 61 7--628 Marras W S, Lavender S A, Leurgans S E et al 1995 Biomechanical risk factors for occupationally related low back disorders. Ergonomics 28: 377-410 Marras W S, Davis K G, Ferguson S A et al 2001 Spine loading characteristics of patients with low back pain compared with asymptomatic individuals. Spine 26(23): 2566-2574 Mirka G A, Kelaher D P 1995 The effects of lifting frequency on the dynamics of lifting. In: Proceedings of the Human Factors and Ergonomics SOCiety, 39th Annual Conference, San Diego, CA, Oct 9-13, Human Factors and Ergonomics Society Morris J M, Lucas D B, Bresler B 1961 Role of the trunk in stability of the spine. The Journal of Bone and Joint Surgery 43A(3): 327-351 Morrissey S J, Liou Y H 1988 Maximal acceptable weights in load carriage. Ergonomics 31: 21 7-226 Mundt D J, Kelsey J L, Golden A L 1993 An epidemiologic study of non-occupational lifting as a risk factor for herniated lumbar intervertebral disc. Spine 18: 595--602 Nachemson A, Elfstrom G 1970 Intravital dynamic pressure measurements in lumbar discs. Scandinavian Journal of Rehabilitation Medicine l (Suppl. 1): 1-40 Natarajan R N, Andersson G B J 1999 The influence of disc height and cross sectional area on the mechanical response of the disc to physiologic loading. Spine 24(28): 1873-1881 Nussbaum M A, Caffin D B, Baker G 1997 Effects of pacing on spinal loads when using material handling systems. In: Proceedings of the Human Factors and Ergonomics Society, 41st Annual Conference, Albuquerque, NM, Sept 22-26, Human Factors and Ergonomics Society Oborne D J 1995 Ergonomics at work. John Wiley, Chichester Rodgers S H J, Yates J W, Garg A 1991 The physiological basis for manual guidelines. National Technical Information Service, report no. 91 -227-330 Schipplein 0 D, Reinsel T E, Andersson G B J et al 1995 The influence of initial horizontal weight placement on the loads at the lumbar spine while lifting. Spine 20(17): 1895-1898 Scholz J P 1993 Organisational principles for the coordination of lifting. Human Movement Science 12: 427-459 Snook S H, Ciriello V M 1991 The design of manual handling tasks: revised tables of maximum acceptable weights and forces. Ergonomics 34(9): 1197-1213 Solomonow M, Zhou B, Harris M, et al 1998 The ligamento-muscular stabilising system of the spine. Spine 23(23): 2552-2562 Solomonow M, Zhou B, Baratta R V et al 1999 Biomechanics of increased exposure to lumbar injury caused by cyclic loading. 1 : Loss of reflexive muscular stabilization Spine 24(23): 2426-2434 Sparto P J, Parnianpour M 1998 Estimation of trunk muscle forces and spinal loads during fatiguing repetitive trunk exertions. Spine 23(23): 2563-2573 Straker L M, Stevenson M G, Twomey L T, Smith L M 1997 A comparison of single and combination manual handling tasks risk assessment: Part 3 Biomechanical measures. Ergonomics 40: 708-728 Straker L, Duncan P 2000 Psychophysical and psychosocial comparison of squat and stoop lifting by young females. Australian Journal of Physiotherapy 46(1): 27-32 Stubbs D A 1985 Human constraints on manual working capacity: effects of age on intra truncal pressures. Ergonomics 28: 107-114 Taimela S, Kankaanpaa M, Luoto S 1999 The effect of lumbar fatigue on the ability to sense a change in lumbar position, a controlled study. Spine 24(13): 1322-1327 Tichauer E R 1971 A pilot study of the biomechanics of sitting in simulated industrial work situations. Journal of Safety Research 3(3): 98-115. Tveit P, Daggfelt K, Hetland S, Thorstensson A 1994 Erector spinae lever arm length variations with changes in spinal curvature. Spine 19(2): 199-204 van der Burg J C E, van Dieen J H 2001 Underestimation of object mass in lifting does not increase the load on the low back. Journal of Biomechanics 34: 1447-1453 van Dieen J H, Kingma I 1999 Total trunk muscle force and spinal compression are lower in asymmetric moments as compared to pure extension moments. Journal of Biomechanics 32: 681--687 van Dieen J H, Hoozemans M J M, Tousaaint H M 1999 Stoop or squat: a review of biomechanical studies on lifting techniques. Clinical Biomechanics 14(10): 685--696 Videman T, Battie M C 1999 The influence of occupation on lumbar degeneration. Spine 124(11): 1164-1168 Waters T R, Putz-Andersson V, Garg A 1994 Applications manual for the revised NIOSH lifting equation. Department of Health and Human Services (NIOSH) publication no. 94-110, National Institute for Occupational Safety and Health, Washington DC Waters T R, Putz Anderson V 1996 Manual materials handling. In: Bhattacharya A, McGlothin J D (eds) Occupational ergonomics. Marcel Dekker, New York, pp 329-350 Yang K H, King A L 1984 Mechanism of facet load transmission as a hypothesis for low back pain. Spine 9: 557-565




Motor control of the cervical spine
E. A. Kesh ner

CHAPTER CONTENTS The head-neck system spine


The problem of motor control in the cervical

105 Musculoskeletal redundancy 106 Vertebral column 106 Muscles of the head and neck 106 Neck muscle organization 107 Neck muscle morphometry 107 Functional synergies 107 Directional patterns of activation 108 Directional tuning of the neck muscles 108 Relationship between neck muscle activation and muscle mechanics 109 Influence of posture on neck muscle activation patterns 109 Changes in cervical spine position 109 Changes in whole body posture 110 Altering the system mechanics 110 Neural control of the cervical spine 110 Characteristics of the cervical spine reflexes 111 Angular vestibulocollic reflex 111 Cervicocollic reflex 111 Linear vestibulocollic reflex 112 Multimodal control of the head and neck 112 Mechanisms controlling head stabilization 112 Influence of task on neck muscle activation patterns 113 114 Models of the head and neck Conclusion 115

Action of the cervical spine cannot be considered as the sum of isolated movements about several joints. It is a dynamic structure which acts to support the head on the trunk, orient the head in space and transmit forces arising from the trunk that will influence the position of the head. Thus, although neural control of the spine operates through the biomechanical structures, the control operations are very much dependent upon the goal of the movement. The musculoskeletal anatomy of the cervical spine has been studied in detail (see, for example, Kamibayashi & Richmond 1998, Sherk & Parke 1983, Worth 1994), but the neurophysiological control of that anatomy by the central nervous system (eNS) has been less focused upon. It is par­ ticularly difficult to identify sources of eNS control in the cervical spine of intact individuals because the presence of vital contents such as the carotid and vertebral arteries, spinal cord and larynx makes invasive recording proce­ dures undesirable. Another reason is that the cervical spine is not analogous to the rest of the vertebral column. The neck is designed for enormous degrees of mobility and the vertebrae are smaller and more anatomically complicated (Bland & Boushey 1988). Finally, any control of the cervical spine must be considered in the context of the redundancies inherent in both its musculoskeletal and sensorimotor com­ ponents.

The cervical spine is a complex biomechanical linkage com­ posed of multiple degrees of freedom of movement about each of its joints and at least 20 pairs of muscles, many of which are capable of performing similar actions. In fact, there appear to be more muscles than are necessary to per­ form the repertoire of head movements that humans make (Peterson et al 1989). This would increase the degrees of freedom for each task by increasing the choice of potential motor patterns available to the eNS. Thus the ultimate



motor control problem in the cervical spine is how to sim­ plify or reduce the degrees of freedom for efficient and timely production of an optimal movement pattern (Bernstein 1967). The purpose of this chapter is to report what is currently known about sensorimotor control of the musculoskeletal and neurophysiological components of the cervical spine in order to examine how the CNS solves the problem of redundancy.
Musculoskeletal redundancy

Complex multiple muscle systems have the potential for pro­ ducing single movements with variable muscle activation patterns. Neural and mechanical redundancies in the head-neck complex potentially provide great flexibility for producing head and neck movements. Longer neck muscles cross many cervical vertebrae and can generate moments about both lower and upper cervical joints. Overall, the num­ ber of independently controlled muscle elements (including subdivisions of compartmentalized muscles) exceeds the number of degrees of freedom of neck motion. Because of their multiple insertions, many of the neck muscles have multiple functions or change their function depending on the initial position of each vertebral joint and the degree to which the joints are free to move in each of the planes of motion (Richmond et al 199 1, 1992, Wickland et aI199 1). The extent of functional variability in the neck muscles appears to depend upon the task being studied. Thus a great deal is known about the static organization of the human cervical spine, but the dynamic organization is less easily character­ ized and data from other species must be relied upon to draw conclusions about control of the cervical spine.
Vertebral column

Qualitative analyses in alert animals demonstrate that most active head movements are made about a restricted set of joint axes (Choi et al 2000, Keshner 199 4). Horizontal plane motions occur about Cl-2, sagittal plane actions about either C1-skull or C7 -Tl and lateral plane motions by combined lateral flexion of C2-5. Fore-aft translation of the head is effected by combining pitch flexion about C7 -Tl with pitch extension about C1-skull to keep the head level. However, all the cervical joints are actually involved to some extent in the final head movement, and the relative amount and pattern of motion at each cervical joint during head movement appears to depend upon the posture that the animal assumes. For example, vertebrae in a standing cat were observed to travel in a vertical, arcing motion, whereas in a prone cat those same vertebrae travelled more diagonally (Keshner 1994). It was determined that the prone cat moved primarily at C1 and C7 with additional motion at C4. Standing cats, however, tended to use all of the vertebrae to accomplish the full excursion of the head (Table 9 .1). Similarly, a rhesus monkey used all of the cervi­ cal vertebrae when tracking a target in the upright position, but moved primarily at C 1-skull when tracking in quadruped. Different muscles were activated to accomplish the same head movements depending upon the position of the animal. The moment arms of the muscles were not sig­ nificantly different in the altered body positions (Keshner et al 1999), therefore different muscles were probably selected by each animal because the trajectory of the spinal column varied for the two body positions.
Muscles of the head and neck

The neurally controlled actions of the large number of mus­ cles of the head and neck must be matched to and interact with the mechanics of the spine. All mammalian species examined hold the cervical vertebral column nearly vertical at rest (Graf et al 1995, Vidal et aI1986). In the cervical spine there are eight joints between the skull and the thoracic ver­ tebrae, each having six degrees of freedom: three rotational (flexion/ extension, axial rotation, and lateral bending) and three translational (up-down, side to side, anterior-poste­ rior). But when analysed descriptively with videofluo­ roscopy, head-neck biomechanics appear much simpler than expected. Videofluoroscopy studies of the cervical spine performed on quadrupedal mammals in a resting position (de Waele et al 1989, Graf et al 1995, Vidal et al 1986) suggest that there are only two primary axes of motion in the cervical spine: the atlanto-occipital joint (C1-skull) and the cervciothoracic joint (C7 -Tl). In human cadavers (Panjabi et a12001), the greatest degree of flexion occurred at CI-C2 and the greatest degree of extension was observed at skull-Cl. Axial rotation was largest at Cl-2 and with lateral bending moments the motion occurred across all vertebral levels.

Although the head represents only 7% of the body's total weight (Gowitzke & Milner 1980), more than 20 different muscles directly link the skull on either side of midline to the vertebral skeleton (Sherk & Parke 1983). Thus, a volun­ tary motor task in the head and neck could be accomplished through a variety of combinations of kinematic and muscle actions. The multiple choices might not be so surprising if

Table 9,1 Mean (± S.D.) angular excursion of each vertebra with respect to T, during ±20' head tracking movements in the sagittal plane in two body postures
Monkey Upright Quadraped Standing Cat Prone

Skull C1 C2 C3 C4 CS C6 C7

32'±4' 27'±3' 25'±5' 22'±4' 24'±1r 17'±2' 16'±4' S'±5'

26'±5' 15'±6' ,,'±3 ' 11'±3 ' 12'±4' S'±3' 7'±2' S'±l'

14'±2' 15'±2' 15'±3' 17'±3' 2 0'±4' 24'±5' 35'±7'

25'±9' 3D'±9' 42'±10' 69'±33'

"Data not available.

Motor control of the cervical spine


the head were involved in the fine motor control and variety of motions found in the hand and fingers. Motions of the head Felative to the trunk, however, are primarily directed towards orienting and stabilizing the position of the eyes and head in space (Goldberg & Peterson 1 986, Outerbridge & Melvill Jones 197 1), even during fine motor activities such as eating and scanning the environment. Despite this oppor­ tunity for response variability, our previous work in both humans (Keshner et al 1989) and cats (Keshner et al 1992, 1997) suggests that, within an animal, the eNS programmes neck muscles to respond in specific directions rather than generating an infinite variety of muscle patterns.
Neck muscle organization

Neck muscles are characteristically grouped in layers. The outer layer consists of long muscles that connect the skull and shoulder girdle. A deeper layer links the skull and ver­ tebral column. The deepest layers consist of muscles that link the cervical and thoracic vertebrae. The outermost layer of muscles, connecting the skull with the shoulder gir­ dle, consists of sternocleidomastoid and trapezius. Sternocleiciomastoid is activated during flexion, contralat­ eral rotation and lateral bending movements of the head. Trapezius has three segments and is classically described as a scapular muscle. Trapezius has been implicated in head extension because the superior fibres of that muscle origi­ nate on the external occipital protuberance (Lockhart et al 1972), although anatomical examination reveals that the muscle fibres become very sparse and essentially disappear around the level of the fourth cervical vertebra. Physiological studies of trapezius do not support its partic­ ipation in pure head extension movements (Keshner et al 1989, Vasavada 1999 , Vasavada et aI 1998). The next layer of muscles links the skull with the vertebral column. This group includes long dorsal (splenius capitis, semispinalis capitis and longissimus capitis) and a long ven­ tral (longus capitis) muscles. Splenius capitis and longis­ simus capitis act as extensors and lateral flexors of the head. Semispinalis capitis is a head extensor muscle that partici­ pates minimally during ipsilateral lateral flexion. Longus capitis lies close to the vertebral bodies, thus it has little mechanical advantage for flexion. Rather it may be activated during ipsilateral rotation of the head. Slightly deeper to these muscles lie splenius cervicis, semispinalis cervicis, longissimus cervicis and longus colli which link the verte­ brae to one another. The functions of these muscles should be similar to those of the capitis segments, but they have smaller moment arms because they lie closer to the vertebrae. The deepest layers of muscles are the suboccipital muscles, which include rectus capitis posterior major and minor, obliquus capitis superior and obliquus capitis inferior. All four mus­ cles produce extension at the atlanto-occipital joint.
Neck muscle morphometry

However, an examination of muscle morphometry reveals considerable heterogeneity that could influence the moment generating capacity and, therefore, the optimal performance of each muscle in each task. In cats, each mus­ cle differs in its relative content of fast and slow fibres, its sites of origin and insertion and the mechanics of action across the individual joints of the cervical column (Richmond & Abrahams 1975, Richmond & Vidal 1988, Richmond et al 199 1, Selbie et al 1993, Wickland et aI1991). Kamibayashi & Richmond ( 1998) studied the neck mus­ cle architecture and morphometry, measuring musculoten­ don, fascicle and sarcomere length as well as muscle mass and pennation angles in ten human cadavers. Just as in cats, human neck muscles were found to be architecturally com­ plex. Many muscles crossed two or more joints and had multiple attachments to different vertebrae. The number of tendons and vertebral level of attachments varied across specimens. Unlike limb muscles that have distinct tendi­ nous attachments to bone, many neck muscles had very lit­ tle tendon at their ends. Instead of distinct tendons, many neck muscles were found to have a complex architecture of internal tendons and aponeuroses. All of these factors affect the biomechanical consequences of muscle action and could influence their selection by the eNS for performance of a task.

From their anatomical descriptions, it would appear that many of the neck muscles perform overlapping actions.

The task for the head-neck controller is to select which of the many muscles and joints will be operative in any move­ ment to meet apparently conflicting goals of stability and mobility. These include producing a stable base of support for the mass of the head and the head's special sensory receptors to keep them in line with a target while permit­ ting full range of motion of the head on the trunk. The prob­ lem for the controller is that these two goals require a decrease in the active degrees of freedom to accomplish joint stability, yet rely on joint flexibility to accomplish the movement. One solution to the degrees of freedom problem is to organize the muscles as functional synergies. A synergy implies a consistent grouping of a set of muscles to accom­ plish a defined task (Tuller et al 1982). Synergies are concep­ tualized as units of control incorporating the muscles around a joint that will act together in a functional fashion. Thus the eNS relies on synergic mechanisms that are composed of a group of muscles that can span several joints and are con­ strained to act as a single unit. Instead of controlling each individual muscle, the eNS need only trigger a synergic unit to produce the joint torques required for a movement (Buchanan et al 1989). In the neck, Richmond et al ( 1992) observed dissociation between deep and superficial neck muscles in freely moving cats that would suggest control by different neural substrates. This differential control might be indicative of independent muscle synergies, each acting at a different axis of joint motion to produce a different



movement goal. Separate actions of two groups of muscles, one producing the forces necessary to move the vertebral col­ umn and the other to align the head with a terminal target, would assist the head-neck controller in meeting its multiple criteria or goals (Thomson et al 1994 ).
Directional patterns of activation


Originally, synergists referred to those muscles producing the same direction of force ( eevor 1977). Experiments on B complex motor systems, including the oral-facial system (Abbs & Cole 1988), the upper arm ( uchanan et al 1989) B and the head and neck (Keshner et al 1989), have demon­ strated that a single action can be accomplished through a variety of muscle patterns consisting of both agonists and antagonists. The controlled parameter appears to be the required force vector rather than the specific force-lever arm of any one muscle (Macpherson 1988, 199 1). It is theoretically possible that each individual head movement is produced by a variety of muscle patterns, thus requiring many muscles to satisfy all possible combi­ nations. In that case, the CNS could then choose from a number of possible combinations to produce the desired outcome (Crowninshield &Brand 198 1). It has been seen in both decerebrate and alert cats that the maximal excitation of a muscle when participating in any particular task is strongly related to a specific direction of motion (Keshner et al 1992, 1997). Thus muscles may be selected on the basis of the required force vector rather than on the goal of the task. Concomitant studies in humans have been impeded by the methodology required to isolate individual neck muscle responses. Experimenters have mostly relied upon the implantation of fine wire electrodes (Takebe et al 1974, Vitti et al 1973 believed to be the only method that could dis­ ), tinguish between the overlapping neck muscles. There is a reluctance on the part of human subjects, however, to undergo invasive procedures with needle or wire elec­ trodes, and these studies have only been able to measure the response of one muscle at any given time. Measurement of a single muscle during any given head movement limits the conclusions that can be drawn con­ cerning the programming of synergistic patterns of action. For example, Zangemeister et al (1982 recorded from two ) pairs of neck muscles with surface electrodes during ballis­ tic head rotations and found that initial head position strongly influenced their results on the functional interac­ tion of the two muscles. Keshner et al (1989) simultaneously recorded the activation of four neck muscles using surface electrodes during an isometric head stabilization task. When amplitude of electromyographic ( MG) activation E was plotted against direction of applied force, each muscle exhibited a preferred direction of activation, which could be represented as a three-dimensional maximal activation direction vector having components in the flexion/ exten­ sion, lateral bending and axial rotation directions (Fig. 9.1). Increasing the force applied to the head linearly increased

lateral flexion


lateral flexion




Semispinalis capitis Splenius capitis Sternocleidomastoid


Figure 9.1 Mean percentage of maximum EMG activation in the frontal plane for three neck muscles in 15 subjects attempting to isometrically stabilize their heads against a 4.5 kg weight demon­ strates overlapping activation but differing maximal activation directions. The point closest to the circumference is the direction of maximal activation for that muscle.

EMG output in preferred response directions and revealed a non-linear change in EMG activation in non-preferred directions, which changed from deactivation at low forces to activation at higher forces. These results demonstrate that describing the combined activation patterns of the muscles during each controlled head movement is neces­ sary to reveal how the CNS programmes this complex motor system.
Directional tuning of neck muscles

Most studies of neck muscle activation examined moments about only one anatomical axis. However, most neck muscles have moment arms for more than one axis of rotation and their resultant moment arm direction is not aligned with an anatomical axis Vasavada (1999) examined neck muscleEMG tuning curves while subjects generated three-dimensional isometric force moments against a stationary six-degrees-of­ freedom rotational load cell attached to a head brace. Force moment output was displayed by a cursor on a computer screen to provide feedback to the subject. This task, where the head was rigidly stabilized, eliminated the need to protect the system against unwanted motion and resulted in more con­ sistent patterns of muscle activation. The neck muscles exhibited significant spatial tuning when subjects generated torques in the flexion, extension, lateral bending and intermediate directions. A test showed that each muscle's preferred direction was unique and con­ sistent among subjects. Preferred activation dfrections were

Motor control of the cervical spine


flexion for sternocleidomastoid, ipsilateral lateral flexion for splenius capitis and extension for semispinalis capitis. Trapezius was tuned toward lateral flexion but had the low­ est activation levels and greatest variability. The maximal activation direction did not always correspond to the direc­ tion of the force moment produced by the muscle. For example, maximal activation of sternocleidomastoid was almost orthogonal to its moment arm direction, which was maximal in lateral flexion. When axial rotation was included in the target moment, each muscle had a unique preferred direction which was consistent among subjects and usually dominated by a strong axial rotation component, shifting the maximal acti­ vation vectors even further from the moment arm direction. This behaviour can be explained by the fact that there are no muscles that produce powerful axial torques and there­ fore the CNS must activate the muscles that are available quite strongly to attain isometric torques equivalent to those in flexion/ extension or lateral bending. Thus, when the head-neck system is used in a situation with a single well-defined goal, neck muscles are activated in a consistent pattern over time and across subjects. When the task has multiple requirements such as maintaining head orientation and protecting the system from excessive forces, multiple strategies appear, each of which may use a different synergistic combination of muscles to produce the required net force moment (Peterson et aI2001).
Relationship between neck muscle activation and muscle mechanics

tial. In other cases, however, changes in muscle activation patterns were observed without changes in muscle moment arms or force generating potential. Thus, the moment gen­ erating potential of the muscles appears to be just one of the variables that influence which muscles the CNS will select to participate in a movement.
Influence of posture on neck muscle activation patterns
Changes in cervical spine position

A likely control parameter for the CNS to employ in select­ ing the muscles that should participate in any given action would be the maximum moment arm of each muscle in order to maintain maximum mechanical advantage for each muscle. In one study (Keshner et a1 1989) muscle tun­ ing often corresponded to the direction of maximum mechanical advantage. Head position was not controlled in this study, and muscle activation patterns were found to differ across subjects, particularly in splenius capitis. When head position was controlled (Mayoux-Benhamou & Revel 1993, Vasavada 1999), directional tuning curves were con­ sistent across subjects but did not correspond to the direc­ tion of the maximum moment arm. The EMG/ moment relationship was non-linear and the maximum extension moment of the dorsal neck muscles occurred in the neutral head position. The relationship between neck muscle activation pat­ terns and maximum moment arm was also examined in cats during a head tracking task in the sagittal plane (Keshner et al 1997). A three-dimensional biomechanical model (Statler 2001) was developed to estimate how muscle moment arms and force generating capacities change dur­ ing the head tracking movement. In some cases, modifica­ tion of muscle activation patterns was consistent with changes in muscle moment arms or force generating poten-

Muscle activation patterns have been found to differ between isometric (load control) and isoinertial (position control) tasks (Buchanan & Lloyd 1995, Tax et al 1990). In cats, spatial activation patterns and temporal relations between each muscle and a tracking device were affected by the plane of motion in which the head movement occurred. The plane of motion also affected the magnitude of muscle activation. It may be that a muscle's length and pulling direction, which can vary with cervical spine orien­ tation, have a greater influence on its contribution to an action than mechanical efficiency (Mayoux-Benhamou & Revel 1993, Runciman & Richmond 1997). Deep and superficial muscles in the cat neck were recorded during head turning movements (Thomson et al 1994). Some muscles varied their activity levels when the neck was horizontally or vertically positioned. Other mus­ cles exhibited the same activity in either position, leading these investigators to purport the presence of both a vary­ ing and invariant synergy in the neck. The two groups of muscles seemed to be separated by those that were more superficial and attached to the lambdoidal crest (invariant) versus those that were lateral and caudal and attached intervertebrally or onto the scapula (variable). The variable synergy was assumed to reflect the changing lever arms of the muscles with changes in body position. It is not clear whether postural effects are due to CNS selection of muscle activation patterns as a result of some fixed synergy, or whether the activation patterns are gov­ erned by the demands of the mechanical task including the mass of the system and the position of the thorax relative to the head (Richmond et aI 1992). Orientation of the cervical spine (i.e. perpendicular or parallel to earth horizontal) was a significant variable in determining both the range of cer­ vical joint motion and the amplitude and timing of the neck muscles (Statler & Keshner 2003). Changing the orientation of the neck did not lead to a large change in the moment arms of any of the muscles examined. Differences in muscle moment arms that did occur were too small to account for the different muscle activation patterns. It would appear that the functional capacity of the muscles was not com­ promised by the small changes in head-neck position required. Thus the mechanical properties of the muscles were not the relevant parameters generating the switch between muscle activation patterns. Differences in EMG activation with initial orientation could be due to a shift in

1 10


the gravitational force vectors on the mechanical system requiring additional energy from the muscles to maintain a common output.
Changes in whole body posture

Horizontal neck
Lag extension

Neck muscle activation was also examined in a rhesus monkey (Macaca mulatta) in two body positions (Choi et al 2000, Peterson et a12001). The animal produced sinusoidal (0.25 Hz) head tracking movements in the sagittal plane when seated with trunk and head vertical or while standing in the quadrupedal position. Vertebral motion was found to vary with body posture, occurring synchronously at all joints in upright, and primarily at skull-C1 in the quad­ rupedal position. EMG activation increased in the quadrupedal position and extensor muscles were concur­ rently activated when the neck was reaching peak exten­ sion. When upright, activation of muscles attaching to the upper cervical column (biventer cervicis, complexus, obliquus capitis inferior, rectus capitis posterior minor) peaked prior to full extension of the neck. Activation of muscles attaching to the lower vertebrae or scapula (rhom­ boideus capitis, semispinalis cervicis, splenius capitis, leva­ tor scapula anterior) peaked when the neck moved towards flexion. Only rectus capitis posterior major responded when the neck was fully extended. Thus, when upright, muscles are activated in functional groupings defined by their anatomical attachments. In the quadrupedal position, gravity acting on the horizontally oriented head produced greater activation and a collective response of the muscles. These results suggest a connection between central recruit­ ment and the mechanical requirements of the task, which might include orientation of the spine relative to gravity and its interaction with muscle length-tension properties (Mayoux-Benhamou et al 1997).
Altering the system mechanics

Neck flexion

0 \l o.


Neck extension


+Og Lead extension

+ 200g Biventer cervicis Complexus Rectus caprtis major Occipitoscapularis Splenius capitis

o o

• • • X T

Vertical neck
Lag extension

o + \l

Neck flexion

r---�&-"'t---If-'--IB.J extension


Lead extension

Figure 9.2 Polar plots of the amplitude (distance from the origin) and phase (location on axis) of the muscle EMG response in the cat during ±20' voluntary head tracking in the sagittal plane at 0.25 Hz. The cervical spine was parallel to the earth horizontal in the top plot and perpendicular to earth horizontal in the bottom plot. Open symbols are the response of the muscle with no weight added to the head. Filled symbols are the response of each muscle when a 200 g weight was placed on the head. Note the greater difference in muscle activation between the two neck postures rather than between the 0 g and 200 g responses.

Changing the inertial mass of the system has less impact on the selection of muscles and the behavioural response than does changing postural orientation (Fig. 9.2). In the cat, adding a weight to the head during a voluntary tracking task had little effect other than to increase the amplitude of each muscle's EMG response (Statler & Keshner 2003). Increased response amplitudes were observed in the sagit­ tal plane when the head was weighted but not in the hori­ zontal plane. In a study of human subjects attempting to stabilize a weighted head during whole body rotations in the horizontal and sagittal planes (Keshner et al 1999), the behaviour of the head was found to change relatively little with added inertia. As adding inertia to a passive mechan­ ical system should cause substantial changes in dynamics, neural mechanisms must be invoked to maintain the con­ stant response dynamics. The principal effect on muscle EMG responses was to increase the delay in the response at high frequencies. Opposing neck muscles continued to be reciprocally activated even with an increased inertial mass.

A mathematical model of head-neck control (Peng et al 1996) has identified stiffness and the vestibulocollic reflex as the primary contributors to the control of head stabiliza­ tion in space. It seems likely that the amount of both stiff­ ness and reflex output are under central control and that these parameters were readjusted to maintain a consistent frequency response pattern of head movement when the head was weighted. Stiffness modulation should play a more significant role in the vertical than in the horizontal plane because of the greater influence of gravity on motions in the vertical plane that stimulates muscle receptors, thereby altering muscle preactivation levels and initial muscle stiffness.

Several areas of the nervous system have been implicated in control of the cervical spine. These can be cond�nsed into

Motor control of the cervical spine


those that initiate the vestibular and cervical proprioceptive reflexes and those involved in voluntary head motions. Reflex, responses of the neck, which include the vestibulo­ collie (VCR) and cervicocollic (CCR) reflexes, respond reac­ tively to acceleratory and proprioceptive stimuli to maintain the orientation of the head in space (VCR) and the head on the trunk (CCR). Voluntary responses are those used for tracking and acquiring exteroceptive (visual, audi­ tory, olfactory) information and can occur as either antici­ patory or pursuit actions. The normal repertoire of head movement responses emerges from combinations of input and output signals. Vestibulospinal pathways are considered to be the pri­ mary conveyor of descending inputs to the neck because of their monosynaptic and disynaptic connections with cervi­ cal motoneurons. Cervical proprioceptive inputs have a sig­ nificant influence on the vestibulospinal signals (Gdowski & McCrea 2000) and a role in head stabilization as evi­ denced in studies of cervical proprioceptive disorders that have demonstrated distinct orientation and postural distur­ bances (Brandt 1996, Karlberg et aI1995). Other descending pathways, particularly those from the reticulospinal system (Peterson 1984, Peterson et a11978), have been shown to be equally important for eliciting the VCR response. Convergence of vestibular and somatosensory input onto the vestibulospinal and reticulospinal (Brink et al 1980) neurons is well documented and occurs at the level of the vestibular nuclear complex and adjacent reticular formation and upon spinal interneurons (Boyle & Pompeiano 1980, Kasper et a11988, Peterson 1984, Wilson et al 199 0). Convergence of afferent input may occur at the level of the motoneuron as well (Brink et aI1980). The combined signals initiate a series of interspinal reflexes that align the body segments (Roberts 1973, Wilson et al 1984). Precise descriptions of the anatomy and neurophysiol­ ogy of the pathways involved in these actions are compre­ hensively described elsewhere (Berthoz et a11992, Wilson & Melvill Jones 1979). However, descriptions of isolated path­ way locations and actions cannot convey how all of these control pathways operate through the biomechanics of the system to produce smooth, purposeful motions of the head and neck. This chapter will focus on the behavioural corre­ lates that elucidate the motor control properties of the head and neck.
Characteristics of the cervical spine reflexes

the head on the body and can provide information about the rotation of the head on the trunk.
Angular vestibulocollic reflex

Specific influences of labyrinthine signals on head and limb movements have been studied in the cat and monkey. With head position fixed relative to the body, and the semicircu­ lar canals stimulated by either angular rotation in the hori­ zontal plane or by electrical stimulation applied to individual canal nerves, a compensatory action of the neck and eye muscles away from the direction of stimulation was evoked (Suzuki & Cohen 1964). Electrical stimulation also results in reciprocal limb movements so that the body is pushed into a vertical position by extension of the oppo­ site limb (Cohen et aI1966). Thus, when the restrained ani­ mal does not have to compensate for normal environmental forces, converging vestibular and spinal inputs either sum or cancel their effects so that the animal appears to attain a position of optimum stability and orientation in space. It is probable that the disynaptic vestibular pathways contribute to the VCR. Nevertheless, the evidence that is available demonstrates that these pathways by themselves are not sufficient to produce the dynamics of the reflex (Wilson & Schor 1999). Other inputs to the VCR include the reticulospinal pathways, which have been shown to replace the short-latency connections when vestibulospinal path­ ways are interrupted (Peterson 1984, Peterson et al 1978). Furthermore, the VCR is not sufficient for purposeful head stabilization in a dynamic environment. The simple response pathways identified in reduced preparations, such as decerebrate cats, are not as readily elicited in alert animals (Banovetz et al 1995, Boyle et al 1996, Wilson & Schor 1999) and are not adequate to produce the forces nec­ essary to counteract external disturbances.
Cervicocollic reflex

The CCR and VCR reflexes appear perfectly suited through their dynamic and somatotopic characteristics to compen­ sate for positional disturbances of the head and neck with respect to the trunk (Dutia & Price 1987, Peterson et a11985, Schor et al 1988). The VCR produces a counter-rotation of the head on the trunk during transient postural reactions that is disturbed in patients with labyrinthine deficit (Horak et al 1994, Keshner et al 1987). The CCR stabilizes

Recent findings (Gdowski & McCrea 2000) suggest that cor­ rect alignment of the head with the trunk and with the gravitoinertial vertical (Imai et al 2001) requires that the vestibular system receive ascending somatosensory inputs. Thus, to attain an appropriate postural response, a conver­ gence of sensory information from the vestibular and somatosensory systems is needed to align the body with respect to earth vertical and to align various body parts with respect to each other. The CCR arises from a stretch of the neck muscles (Ezure et al 1983, Peterson et al 1985) as would occur if the body were turning under a fixed head. Although there is an abundance of evidence that the neck proprioceptors pro­ vide information regarding the position and movement of the head with respect to the trunk, the actual receptors elic­ iting this response are not well delineated. The muscle spin­ dle is the most obvious receptor to produce the CCR response, but the physiological evidence suggests this reac­ tion is not a simple spindle reflex (Wilson 1992). The evi­ dence suggests the presence of more signal processing


between the receptors and motoneurons than expected from a monosynaptic pathway, and there is evidence for a contribution from presynaptic inhibition in the CCR response (Banovetz et al 1995, Wilson 1988).
Linear vestibulocollic reflex

.1 1

Frequency (Hz) : f\




Because of the difficulty in isolating outputs from the otoliths, little is known about the actions of this component of the vestibular system. The otoliths are stimulated by lin­ ear accelerations of the head and their inputs have been found to modify both eye and head stabilizing responses (Schor & Miller 198 1 , 1982, Schor et aI 1985). Studies of neck reflex activation in alert cats (Banovetz et al 1 995, Lacour et al 1987) and head stabilization in humans (Keshner et al 1995) found VCR responses to pitch or roll rotations that contain an added component from activation of otolith organs. Otolith contributions to compensatory eye and neck responses increased with stimulus frequency, but the otolith system alone was unable to produce perfect com­ pensation (Borel & Lacour 1982). There is some suggestion that convergence of canal and otolith input on vestibu­ lospinal neurons supplies combined angular and linear acceleration inputs to the vestibulospinal reflexes in the neck (Uchino et al 2000); however, the otoliths have been described as having a distinct functional effect during loco­ motion. In order to maintain a stable head fixation distance over the optimal range of walking velocities, it was pro­ posed that compensatory head pitch movements were produced predominantly by the angular vestibulocollic reflex at low walking speeds and by the linear vestibulocol­ lic reflex at higher speeds ( Hirasaki et aI 1999).

il� <

.--='-:-Frequency (Hz) --,

:� �
.1 1


-,.� : I �p- 50ms ��

Prop,rioceptive Inputs 1---: ::-:-------' �

Frequency (Hz)


I i�
I !\






Figure 9.3 Flow diagram depicting the potential pathways partic­ ipating in head stabilization. Graphs surrounding each control box illustrate the expected activation of each control mechanism in the time (latency) and frequency domains.

humans undergoing horizontal trunk rotations when the trunk was fixed (Bizzi et al 1 976, Gresty 1987, Guitton et al 1986, Keshner & Peterson 1995) and in humans in the verti­ cal plane (Keshner et aI 1995).
Mechanisms controlling head stabilization

Although the CCR and VCR appear to be all that is neces­ sary to compensate for positional disturbances of the head and neck, most of the research indicates that more than one mechanism actually contributes to control of the cervical spine (Fig. 9.3). Multiple pathways contribute to stabiliza­ tion of the head by the neck, and there is evidence that remaining pathways may compensate for the loss or injury of any one pathway, although it would not be a complete functional substitution (Keshner 2000, Keshner & Chen 1996). In a series of studies in which seated subjects were rotated with predictive sinusoids, the natural mechanics of the head-neck system were found to be adequate to pro­ duce head stabilization (Barnes & Rance 1974, 1975, Bizzi et aI 1 978). Other studies that examined stabilization of the head when the trunk was fixed, however, demonstrated that the VCR and CCR also made a strong contribution to this response (Goldberg & Peterson 1 986, Keshner & Peterson 1995, Viviani & Berthoz 1 975). Voluntary motor commands and mechanical properties of the system also accounted for compensatory head motions in monkeys and

Differential frequency characteristics of the mechanisms potentially controlling head stability ( Fig. 9.4) were revealed through a paradigm of fixing the trunk to the seat so that only the head was free to move. Subjects were ran­ domly rotated in the dark in the vertical and horizontal planes (Keshner & Chen 1 996, Keshner & Peterson 1995, Keshner et a1 1 995) with a range of sinusoidal frequencies. Voluntary control was manipulated by either distracting the subjects with a mental task or requiring that they stabi­ lize to a visual image. In the horizontal plane, responses of the neck with respect to the trunk were similar to responses that were pre­ dicted by an underdamped, second order linear system (see Fig. 9.4). Amplitudes of the neck with respect to trunk amplitudes rose at approximately 40 dB/ decade and, when amplitudes were large enough to be meaningful, response phases shifted from 0 degrees to - 180 degrees. A plateau in the response dynamics that appeared around 1.5-2.5 Hz (see arrow in Fig. 9.4) implies that an additional control mechanism was operating to damp the rising response. Resonant responses of the head (implying that stabilization was lost and the head moved more than the trunk) were observed above 2.5 Hz. These data suggest that the head was locked to the trunk at low frequencies, had a small area


control of the cervica l spine

1 13

VCR 360





-180 ompensati -360 '-----360 01 . 10 0 .1 -180 Frequency (Hz)

/"--... 6 �


Figure 9.4 The cartoon on the left depicts the mechanisms con­ tributing to stabilization of the head. Canal and otolith signals descend from the labyrinths to excite the contralateral and inhibit the ipsilateral neck muscles via the VCR. Proprioceptive inputs from the neck generate excitation of the ipsilateral neck muscles via the CCR. Descending voluntary signals (?!) and the intrinsic mechanics of the neck such as viscoelasticity (B) and stiffness (K) also influ­ ence the response. On the right are bode plots of the amplitude and phase of the average neck with respect to trunk response to hori­ zontal and vertical plane pseudorandom sum-of-sine rotations dur­ ing a mental distraction task. Phases of ±l aO· and amplitudes equal to 1 indicate that the head is perfectly compensating for motion of the trunk to stay stable in space. A phase of O· indicates that the head is moving with the trunk. Amplitudes greater than 1 mean that the head is moving more than the trunk. The vertical arrow indicates the plateau in the horizontal plane dynamics that appeared around 1.5-2.5 Hz.

1 Frequency (Hz)


responses controlled by visual signals. Above 1 Hz, mechanical factors (e.g., inertia, stiffness and viscoelastic­ ity) become important and act with little delay. Linear translations of the subject when the trunk was free to move also reveal a significant somatosensory com­ ponent of head stabilization as a result of motion of the trunk at the neck (Forssberg & Hirschfeld 1994, Gresty 1989, Vibert et al 2001). The relative importance of the VCR and CCR for head-neck stabilization is probably dependent upon the degrees of freedom and postural requirements of the task.
Influence of task on neck muscle activation patterns

of compensation for trunk motion, and then stabilization was lost as frequency increased. The simple second order response pattern would suggest that voluntary mecha­ nisms act to stabilize the head to the trunk at low frequen­ cies and that head inertia and other biomechanical factors predominate at high frequencies. In the mid-frequency range, reflexes may become active to dampen and delay control by the system mechanics. In the vertical plane, larger response amplitudes imply that head stabilization improves at low frequencies « 1 Hz). The response is sim­ ilar to the responses in the horizontal plane at the mid­ frequencies but the high frequency resonance is not observed. Enhanced otolith contributions may have occurred at low frequencies in the vertical plane due to gravitational input, thereby smoothing the performance of the head stabilizing system throughout the frequency range. Thus, the reflexes appear to be predominant in the fre­ quency range of natural locomotion ( 1.5-2 Hz) (Hirasaki et al 1999) and their function is to damp oscillations of the head at higher frequencies or with altered system mechan­ ics (Keshner et al 1999). Below 1 Hz, the head is primarily stabilized by longer latency (>100 ms) voluntary pathways and, perhaps, otolith signals. Voluntary responses are observed as anticipatory torques in the neck muscles or

In alert cats, movements generated in a particular direc­ tion during a voluntary head-tracking task used different muscle patterns than the same head movements gener­ ated by the neck reflexes (Keshner et al 1992). Correspondingly, the maximal response of individual muscles occurred at different orientations for the two tasks (Fig. 9 .5). But each voluntary and reflex head move­ ment in the cat was produced by an identifiable and repeatable pattern of neck muscle activation during ori­ enting and stabilizing behaviours (Baker et al 1985, Keshner et al 1992, Roucoux & Crommelinck 1988). This was also true in head-fixed monkeys during pursuit eye movements (Lestienne et al 1984). This would imply that each head motion task is executed by a specific muscular pattern that is not repeated in any other direction. Different patterns of muscle activation during reflex and voluntary head motions suggest that the sensorimotor transformation process is different for reflex and voluntary

I �


• • •

• S

Orientation angle Orientation angle


Figure 9.5 Plots of amplitudes and phases of the right complexus muscle EMG responses during ±20· voluntary head tracking and VCR trials at 0.25 Hz in different head orientations. Responses are derived from a least-squares fit to five days of data from one cat. A sinusoid fit to the amplitude data i llustrates the sinusoidal pattern of EMG output with maximum and minimum responses shifted +22· in the VCR task. A 90· phase shift in the VCR relative to track­ ing indicates a response related to the velocity rather than the position of the head.

1 14


tasks, thereby modifying the directional results. Numerous sites have the potential to be a locus for the sensorimotor transformation of voluntary movements. Neurons of the pontomedullary reticular formation, many that monosy­ naptically excite motoneurons supplying neck and axial muscles (Peterson et al 1978, 1984), get inputs from head and trunk areas of motor cortex in the cat ( Alstermark et al 1983, Peterson et al 1975). Convergent semicircular canal and neck proprioceptive inputs were recorded at cortical levels in alert cats during a passive rotation task ( Mergner et aI1985). There are also widespread reciprocal projections between cerebellum and neck afferents (Chan et al 1982, Wilson et aI 1976). Transformation of vestibular inputs to neck motor output during the VCR occurs primarily in the brainstern nuclei. Head movements need to be constrained during the reflex task and may include only a few joints, thereby restricting the system to one pattern of muscle activation, whereas motor solutions for voluntary head tracking need constant adjustment. Multiple sensory input is also operative during voluntary movements, as are changing muscle lengths, mul­ tiarticular motions and a changing visual scene.

The head and neck serve as a strong correlate of the whole body during postural restabilization because of their multi­ segmental, multi-muscle arrangement (Graf et al 1997, Winters & Goldsmith 1983). A critical gap in our knowledge is at the output end where we know very little about the biomechanical action of neck muscles as a function of neck geometry. The complexity of the neck motor system poses a difficult challenge for creating useful predictive models. The most common approach to a dynamic model of the head and neck is the lumped parameter model where sin­ gle parameters are used to represent the inertia, viscosity and elasticity of the system. Goldberg & Peterson ( 1986) have shown that the lumped parameter model provides an excellent fit to properties of a passive head-neck system. However, discrepancies between rigid models and physical data exist and suggest a need in the models for greater free­ dom of joint motion. A biomechanical model first developed to study how surgical changes in musculoskeletal geometry and musculo­ tendon parameters affect muscle force and its moment about the joints (Delp & Loan 1995) has been applied to the cat (Keshner et al 1997 , Statler 2001, Statler & Keshner 2003) and to the human (Vasavada 1999, Vasavada et al 1998) neck. The model uses a graphical interface that allows visu­ alization of the musculoskeletal geometry and permits manipulation of the model parameters. To create a model using this system, the geometry of the bones, the kinemat­ ics of the joints and the lines of action and force generating parameters (physiological cross-sectional area, muscle fibre length, tendon slack length and fibre pennation angle) of the muscles are specified. Once musculoskeletal geometry

is specified, muscle lengths and moment arms can be com­ puted over a range of body positions. Given a set of muscle activation patterns from electromyographic recordings, the forces and moments generated by each modelled mus­ cle can be estimated. Also, the moments developed by pas­ sive structures such as intervertebral ligaments can be incorporated. Moment arms of each muscle are computed from the mathematical descriptions of the muscle lines of action and the joint kinematics. The model can be used to predict the motor control consequences occurring as a result of cervical joint limitations. A homeomorphic model of head and neck sensorimotor integration has been developed (Keshner et al 1999 , Peng et a11996) to interpret experimental data from human sub­ jects. The model is 'lumped' parameter in type because of gaps in available data and to avoid unnecessary complex­ ity. The model is based on the biomechanics, that is, the geometry and physics, of the joints and masses involved. Layered on top of the biomechanics are stiffness (position dependence), viscosity (velocity dependence) and extrinsic torques. The goal is to split out contributions of specific sen­ sory loops and motor control pathways that are relevant to human health. The model (Fig. 9.6) simulating the response of the head to a horizontal trunk displacement incorporates head mechanics, the VCR and the CCR, with parameters drawn from numerous experimental studies (Peng et al 1996). A more complex two-joint model of pitch-plane head motion including VCR and CCR loops has also been devel­ oped and can simulate experimental results (Keshner et al 1999), but the addition of the second joint has increased the mechanical complexity. In the pitch plane the head is unsta­ ble without active control. In response to a step input, it

Trunk acceleration

_ ...L.- -I _ _
Head acceleration WRT space

Extrinsic head torque


Desired head


Active torque




Somato (CCR)



Visual acceleration

Figure 9.6 Control loops believed to participate in head stabilization and incorporated into the homeomorphic model of head stability. In addition to the inertial (I), cervicocollic (CCR), and vestibulocollic (VCR) inputs, somatosensory, visual (visuocollic reflex OCR) and vestibular error signals (shown as ± control signals) are combined, delayed and coupled to the head.

Motor control of the cervical spine

1 15

'falls over' with a pronounced 'bounce' on the top trace of the time domain simulation when there is no compensa­ tion. The addition of static vestibular or proprioceptive inputs results in a head that still leans forwards but remains much closer to upright. The addition of dynamic compen­ sation using the VCR and CCR improves stability.

specific muscle synergy that is presumably optimized to efficiently meet the demands of the task and the neural con­ trollers must compensate for these task and posture dependent variations. Models need to be further developed to explain and delineate the multiple levels of control and response in the cervical spine.

Dynamic studies have indicated that visual and voluntary control of neck muscles and the dynamic and static VCR and CCR preferentially govern the head-neck system in different frequency domains. Thus neural control of the cer­ vical system may be redundant but it is not excessive. Each component of the system is necessary to have a flexible and functional system. Redundant control allows the system to compensate for injury as well as creating a potential for substantial variability within and between subjects. Kinematic studies have indicated the existence of specific muscle activation patterns for voluntary force generation in the neck, of reflex and voluntary control strategies for sta­ bilizing the head during body perturbations, and of several control strategies for voluntary head tracking that vary with posture. Each strategy appears to be executed by a

KEYWORDS biomechanical model cervical spine cervicocollic eNS directional tuning electromyography EMG head tracking kinematics mathematical model moment arms muscle activation patterns neck muscles neural control posture redundancy reflex reticulospinal vertebrae vestibular vestibulocol lic vestibu lospinal videofluoroscopy voluntary control

Abbs J H, Cole K J 1988 NelUal mechanisms of motor equivalence and goal achievement. In: Wise S P (ed) Higher brain functions: recent explorations of the brain's emergent properties. Wiley, New York, pp 15-43 Alstermark B, Pinter M, Sasaki S 1983 Convergence on reticulospinal nelUons mediating contralateral pyramidal disynaptic EPSPS to neck motoneurons. Brain Research 259: 151-154 Baker J, Goldberg J, Peterson B 1985 Spatial and temporal response properties of the vestibulocollic reflex in decerebrate cats. Journal of NelUophysiology 54: 735-756 Banovetz J, Baker J F, Peterson B W 1995 Spatial coordination by vestibular descending signals. I: Reflex excitation of neck muscles i n alert and decerebrate cats. Experimental Brain Research 105: 345-362 Barnes G R, Rance B H 1974 Transmission of angular acceleration to the head in the seated human subject. Aerospace Medicine 45: 4121-4126 Barnes G R, Rance B H 1975 Head movement induced by angular oscillation of the body in the pitch and roll axes. Aviation Space Environmental Medicine 46: 987-993 Beevor C E 1977 The Croonian lectures on muscular movements and their representation in the central nervous system. In: Payton 0 D, Hirt S, Newton R A (eds) Scientific bases for nelUophysiologic approaches to therapeutic exercise: an anthology. F A Davis, Philadelphia, pp 5-12 Bernstein N 1967 The problem of interrelation of co-ordination and localization. In: Bernstein N The coordination and regulation of movements. Pergamon Press, New York, pp 15-59 Berthoz A, Graf W, Vidal P P (eds) 1992 The head-neck sensory motor system. Oxford, New York Bizzi E, Polit A, Morasso P 1976 Mechanisms underlying achievement of final head position. Journal of NelUophysiology 39: 435-444 Bizzi E, Dev P, Morasso P, Polit A 1978 Effect of load disturbances dlUing centrally initiated movements. Journal of Neurophysiology 41: 542-556 Bland J H, Boushey D R 1988 The cervical spine, from anatomy and phYSiology to clinical care. In: Berthoz A, Graf W, Vidal P P (eds) The head-neck sensory motor system. Oxford, New York, pp 135-140 Borel L, LacolU M 1982 Functional coupling of the stabilizing eye and head reflexes during horizontal and vertical linear motion in the cat. Experimental Brain Research 91: 191-206 Boyle R, Pompeiano 0 1980 Responses of vestibulospinal neurons to sinusoidal rotation of neck. Journal of NelUophysiology 44: 633- 649 Boyle R, Belton T, McCrea R A 1996 Responses of identified vestibulospinal neurons to voluntary and reflex eye and head movements in the alert squirrel monkey. Annals of the New York Academy of Sciences 781: 244-263 Brandt T 1996 Cervical vertigo: reality or fiction? Audiology and Neuro-Otology 1: 187-196 Brink E E, Hirai N, Wilson V J 1980 Influence of neck afferents on vestibulospinal neurons. Experimental Brain Research 38: 285-292 Buchanan T S, Lloyd D G 1995 Muscle activity is different for humans performing static tasks, which require force control and position control. NelUoscience Letters 194: 61-64 Buchanan T S, Rovai G P, Rymer W Z 1989 Strategies for muscle activation during isometric torque generation at the human elbow. Journal of Neurophysiology 62: 1201-1212 Chan Y S, Manzoni D, Pompeiano 0 1982 Response characteristics of cerebellar dentate and lateral cortex neurons to sinusoidal stimulation of neck and labyrinth receptors. Neuroscience 7: 2993-3011 Choi H, Keshner E A, Peterson B W 2000 Primate neck muscle activation patterns during head tracking in two postures. Society for Neuroscience Abstracts 26: 2212

1 16


Cohen B, Tokumasu K, Goto K 1966 Semicircular canal nerve eye and head movements: the effect of changes in initial eye and head position on the plane of the induced movement. Archives of Ophthalmology 76: 523-531 Crowninshield R D, Brand R A 1981 A physiologically based criteria of muscle force prediction in locomotion. Journal of Biomechanics 14: 793-801 de Waele C, Graf W, Josset P, Vidal P P 1989 A radiological analysis of the postural syndromes following hemilabyrinthectomy and selective canal and otolith lesions in the guinea pig. Experimental Brain Research 77: 1 66--1 82 Delp S L, Loan J P 1995 A graphics-based software system to develop and analyze models of musculoskeletal structures. Computers in Biology and Medicine 25: 21-34 Dutia M B, Price R F 1987 Interaction between the vestibulocollic reflex and the cervicocollic stretch reflex in the decerebrate cat. Journal of Physiology (London) 387: 19-30 Ezure K, Fukushima K, Schor R H, Wilson V J 1983 Compartmentalization of the cervicocollic reflex in cat splenius muscle. Experimental Brain Research 51 : 397-404 Forssberg H, Hirschfeld H 1994 Postural adjustments in sitting humans following external perturbations: muscle activity and kinematics. Experimental Brain Research 97: 515-527 Gdowski G T, McCrea R A 2000 Neck proprioceptive inputs to primate vestibular nucleus neurons. Experimental Brain Research 135: Goldberg 1, Peterson B W 1986 Reflex and mechanical contributions to head stabilization in alert cats. Journal of Neurophysiology 56: 857-875 Gowitzke B A, Milner M 1980 Understanding the scientific basis of human movement. Williams and Wilkins, Baltimore Graf W, De Waele C, Vidal P P 1995 Functional anatomy of the head-neck movement system of quadrupedal and bipedal mammals. Journal of Anatomy 186: 55-74 Graf W, Keshner E, Richmond F J, Shinoda Y, Statler K, Uchino Y 1997 How to construct and move a cat's neck. Journal of Vestibular Research 7: 219-237 Gresty M 1987 Stability of the head: studies in normal subjects and in patients with labyrinthine disease, head tremor, and dystonia. Movement Disorders 2: 165-185 Gresty M 1989 Stability of the head in pitch (neck flexion--extension): studies in normal subjects and patients with axial rigidity. Movement Disorders 4: 233-248 Guitton D, Kearney R E, Wereley N, Peterson B W 1986 Visual, vestibular and voluntary contributions to human head stabilization. Experimental Brain Research 64: 59-69 Hirasaki E, Moore S T, Raphan T, Cohen B 1999 Effects of walking velocity on vertical head and body movements during locomotion. Experimental Brain Research 127: 117-130 Horak F B, Shupert C L, Dietz V, Horstmann G 1994 Vestibular and somatosensory contributions to responses to head and body displacements in stance. Experimental Brain Research 100: 93-106 Imai T, Moore S T, Raphan T, Cohen B 2001 Interaction of the body, head, and eyes during walking and turning. Experimental Brain Research 136: 1-18 Karnibayashi L K, Richmond F J 1998 Morphometry of human neck muscles. Spine 23: 1314-1323 Karlberg M, Persson L, Magnusson M 1995 Impaired postural control in patients with cervico-brachial pain. Acta Oto-Laryngologica 520 (2) (SuppL): 440-442 Kasper J, Schor R H, Wilson V J 1988 Response of vestibular neurons to head rotations in vertical planes. II: Response to neck stimulation and vestibular neck-interaction. Journal of Neurophysiology 60: 1765-1778 Keshner E A 1994 Vertebral orientations and muscle activation patterns during controlled head movements in cats. Experimental Brain Research 98: 546-550 511-526

Keshner E A 2000 Modulating active stiffness affects head stabilizing strategies in young and elderly adults during trunk rotations in the vertical plane. Gait and Posture 1 1 : 1-11 Keshner E A, Chen K J 1996 Mechanisms controlling head stabilization in the elderly during random rotations in the vertical plane. Journal of Motor Behavior 28: 324-336 Keshner E A, Peterson B W 1995 Mechanisms controlling human head stability. I : Head-neck dynamics during random rotations in the horizontal plane. Journal of Neurophysiology 73: 2293-2301 Keshner E A, Allum J H J, Pfaltz C R 1987 Postural coactivation and adaptation in the sway stabilizing responses of normals and patients with bilateral peripheral vestibular deficit. Experimental Brain Research 69: 66--72 Keshner E A, Campbell D, Katz R, Peterson B W 1989 Neck muscle activation patterns in humans during isometric head stabilization. Experimental Brain Research 75: 335-364 Keshner E A, Baker J F, Banovetz J, Peterson B W 1992 Patterns of neck muscle activation in cats during reflex and voluntary head movements. Experimental Brain Research 88: 361-374 Keshner E A, Cromwell R, Peterson B W 1995 Mechanisms controlling human head stability. II: Head-neck characteristics during random rotations in the vertical plane. Journal of Neurophysiology 73: 2302-2312 Keshner E A, Statler K D, Delp S L 1997 Kinematics of the freely moving head and neck in the alert cat. Experimental Brain Research 115: 257-266 Keshner E A, Hain T C, Chen K J 1999 Predicting control mechanisms for human head stabilization by altering the passive mechanics. Journal of Vestibular Research 9: 423-434 Lacour M L, Borel J, Barthelemy J, Harlay S, Xerri C 1987 Dynamic properties of the vertical otolith-neck reflexes in the alert cat. Experimental Brain Research 65: 559-568 Lestienne F, Vidal P P, Berthoz A 1984 Gaze changing behaviour in head restrained monkey. Experimental Brain Research 53: 349-356 Lockhart R D, Hamilton G F, Fyfe F W 1972 Anatomy of the human body. J B Lippincott, New York Macpherson J M 1988 Strategies that simplify the control of quadrupedal stance. I: Forces at the ground .. Journal of Neurophysiology 60: 204-217 Macpherson J M 1991 How flexible are muscle synergies? In: Humphrey D R, Freund H J (eds) Motor control: concepts and issues. Wiley, New York, pp 33-47 Mayoux-Benhamou M A, Revel M 1993 Influence of head position on dorsal neck muscle efficiency. Electromyography and Clinical Neurophysiology 33: 161-166 Mayoux-Benhamou M A, Revel M A, Vallee C 1997 Selective electromyography of dorsal neck muscles in humans. Experimental Brain Research 113: 353-360 Mergner T, Becker W, Deecke L 1985 Canal-neck interaction in vestibular neurons of the cat's cerebral cortex. Experimental Brain Research 61: 94-108 Outerbridge J S, Melvill Jones G 1971 Reflex vestibular control of head movements in man. Aerospace Medicine 42: 935-940 Panjabi M M, Crisco J J, Vasavada A et al 2001 Mechanical properties of the human cervical spine as shown by three-dimensional load­ displacement curves. Spine 26: 2692-2700 Peng G C, Hain T C, Peterson B W 1996 A dynamical model for reflex activated head movements in the horizontal plane. Biological Cybernetics 75: 309-319 Peterson B W 1984 The reticulospinal system and its role in the control of movement. In: Barnes C D (ed) Brainstem control of spinal cord function. Academic Press, New York, pp 27-86 Peterson B W, Maunz R A, Pitts N G, Mackel R G 1975 Patterns ,Iilf projection and branching of reticulospinal neurons. Experimental Brain Research 23: 333-351 Peterson B W, Pitts N G, Fukushima K, Mackel R 1978 Reticulospinal excitation and inhibition of neck motoneurons. Experimental Brain Research 32: 471-489

Motor control of the cervical spine

1 17

Peterson B W, Goldberg J, Bilotto G, Fuller J H 1985 The cervicocollic reflex: its dynamic properties and interaction with vestibular refl�xes. Journal of Neurophysiology 54: 90-109 Peterson B W, Pellionisz A J, Baker J F, Keshner E A 1989 Functional morphology and neural control of neck muscles in mamma ls. American Zoologist 29: 139-149 Peterson B W, Choi H, Hain T, Keshner E, Peng G C 2001 Dynamic and kinematic strategies for head movement control. Annals of the New York Academy of Sciences 942: 381-393 Richmond F J R, Abrahams V C 1975 Morphology and enzyme histochemistry of dorsal muscles of the cat neck. Journal of Neurophysiology 38: 1312-1321 Richmond F J R, Vidal P P 1988 The motor system: joints and muscles of the neck. In: Peterson B W, Richmond F J (eds) Control of head movement. Oxford University Press, New York, p 121 Richmond F J R, Gordon D C, Loeb G E 1991 Heterogeneous structure and function among intervertebral muscles. In: Berthoz A, Graf W, Vidal P P (eds) The head-neck sensory motor system: Oxford University Press, New York, pp 101-103 Richmond F J R, Thomson D B, Lob G E 1992 Electromyographic studies of neck muscles in the intact cat. I: Patterns of recruitment underlying posture and movement during natural behaviors. Experimental Brain Research 88: 41-58 Roberts T D M 1973 Reflex balance. Nature 244: 156-158 Roucoux A, Crommelinck M 1988 Control of head movement during visual orientation. In: Peterson B W, Richmond F J (eds) Control of head movement. Oxford University Press, New York, pp 208-223 Runciman R J, Richmond F J 1997 Shoulder and forelimb orientations and loading in sitting cats: implications for head and shoulder movement. Journal of Biomechanics 30: 911-919 Schor R H, Miller A D 1981 Vestibular reflexes in neck and forelimb muscles evoked by roll tilt. Journal of Neurophysiology 46: 167-178 Schor R H, Miller A D 1982 Relationship of cat vestibular neurons to otolith-spinal reflexes. Experimental Brain Research 47: 137-144 Schor R H, Miller A D, TImerick S J B, Tomko D L 1985 Responses to head tilt in cat central vestibular neurons. II: Frequency dependence of neural response vectors. Journal of Neurophysiology 53: 1444-1452 Schor R H, Kearney R E, Dieringer N 1988 Reflex stabilization of the head. In: Peterson B W, Richmond F J (eds) Control of head movement. Oxford University Press, New York, pp 141-166 Selbie W S, Thomson D B, Richmond F J R 1993 Suboccipital muscles in the cat neck: morphometry and histochemistry of the rectus capitis muscle complex. Journal of Morphology 216: 47-63 Sherk H H, Parke W W 1983 Normal adult anatomy. In: Cervical Spine Research Society (ed) The cervical spine. J B Lippincott, New York, pp 8-22 Statler K D 2001 A computer graphics based model of the cat head and neck used to examine joint movement, moment generating potential and EMG patterns in voluntary head and neck movements. PhD Thesis, Department of Biomedical Engineering, Northwestern University, Evanston, Illinois Statler K D, Keshner E A 2003 Effects of inertial load and cervical-spine orientation on a head tracking task in the alert cat. Experimental Brain Research 148: 202-210. Suzuki J-I, Cohen B 1964 Head, eye, body, and limb movements from semicircular canal nerves. Experimental Neurology 10: 393-406 Takebe K, Vitti M, Basmajian J V 1974 The functions of semispinalis capitis and splenius capitis muscles: an electromyographic study. Anatomical Records 179: 477-480 Tax A A M, Denier van der Gon J J, Erkelens C J 1 990 Differences in central control of m. biceps brachii in movement tasks and force tasks. Experimental Brain Research 79: 138-142

Thomson D B, Loeb G E, Richmond F J R 1994 Effect of neck posture on the activation of feline neck muscles during voluntary head turns. Journal of Neurophysiology 72: 2004-2014 Tuller B, Turvey M T, Fitch H L 1982 The Bernstein perspective. II: The concept of muscle linkage or coord.inative structure. In: Kelso New Jersey, pp 253-270 Uchino Y, Sato H, Kushiro K, Zakir M M, Isu N 2000 Canal and otolith inputs to single vestibular neurons in cats. Archives of Italian Biology 138: 3-13 Vasavada A N, Peterson B W, Delp S L 2002 Three-dimensional spatial tuning of neck muscle activation in humans. Experimental Brain Research 147: 437-448 Vasavada A N, Li S, Delp S L 1998 Influence of muscle morphometry and moment arms on the moment-generating capacity of human neck muscles. Spine 23: 4 1 2-422 Vibert N, MacDougall H G, de Waele C et al 2001 Variability in the control of head movements in seated humans: a link with whiplash injuries. Journal of Physiology 532: 851-868 Vidal P P, Graf W, Berthoz A 1986 The orientation of the cervical vertebral column in unrestrained awake animals. I: Resting position. Experimental Brain Research 6 1 : 549-559 Vitti M, Fujiwara M, Basmajian J V, Iida M 1973 The integrated roles of longus colli and sternocleidomastoid muscles: an electromyographic study. Anatomical Records 177: 471-484 Viviani P, Berthoz A 1 975 Dynamics of the head-neck system in response to small perturbations: analysis and modeling in the frequency domain. Biological Cybernetics 1 9 : 19-37 Wickland C R, Baker J F, Peterson B W 1991 Torque vectors of neck muscles in the cat. Experimental Brain Research 84: 649-659 Wilson V J 1988 Convergence of neck and vestibular signals on spinal interneurons. Progress in Brain Research 76: 137-143 Wilson V J 1992 Physiologic properties and central actions of neck muscle spindles. In: Berthoz A, Graf W, Vidal P P (eds) The head-neck sensory motor system. Oxford University Press, New York, pp 1 75-178 Wilson V J, Melvill Jones G 1979 Mammalian vestibular physiology. Plenum Press, New York. Wilson V J, Schor R H 1 999 The neural substrate of the vestibulocollic reflex. What needs to be learned? Experimental Brain Research 1 29: 483-493 Wilson V J, Maeda M, Franck J I, Shimazu H 1976 Mossy fiber neck and second order labyrinthine projections to cat flocculus. Journal of Neurophysiology 39: 301-310 Wilson V J, Ezure K, T unerick S J B 1984 Tonic neck reflex of the decerebrate cat: response of spinal interneurons to natural stimulation of neck and vestibular receptors. Journal of Neurophysiology 5 1 : 567-577 Wilson V J, Yamagata Y, Yates B J, Schor R H, Nonaka S 1 990 Response of vestibular neurons to head rotations in vertical planes. III: Response of vestibulocollic neurons to vestibular and neck stimulation. Journal of Neurophysiology 164: 1 695-1703 Winters J M, Goldsmith W 1983 Response of an advanced head-neck model to transient loading. Journal of Biomechanical Engineering 105: 63-70, 1 96-197 Worth D R 1994 Movements of the head and neck. In: Boyling J D, Palastanga N, Jull G A, Lee D G (eds) Grieve's Modern Manual Therapy, 2nd edn. Churchill Livingstone, Edinburgh, pp 53-68 Zangemeister W H, Stark L, Meienberg 0, Waite T 1982 Neural control of head rotation: electromyographic evidence. Journal of Neurological Sciences 55: 1-14


(ed) Human motor behavior: an introduction. Lawrence Erlbaum,





Motor control of the trunk
P. W. Hodges

119 119


Biomechanical demands for control of movement and stability Models of stability Control in neutral Control elements Muscles
121 122 123 120 121


Intrinsic lumbopelvic muscles Sensors
123 125 125

Superficial lumbopelvic muscles Controller

Control models

Open-loop control of the trunk Closed-loop control of the trunk Control of muscle stiffness of the trunk trunk
129 129 128

125 127

Integrated control of stability and movement Factors that complicate motor control of the The effect of pain and injury on motor control Changes in open-loop control mechanisms mechanisms
129 130 132 133 129

Changes in closed-loop control Mechanism of changes in motor control Task conflict of the trunk muscles Respiration Continence
133 134 134 134

Other factors leading to task conflict Implications of task conflict Additional control issues Conclusion
134 134

It is well accepted that the spine is inherently unstable and dependent on the contribution of muscles in addition to the passive elements of the spine to maintain stability and to control movement (Panjabi 1992b). Although trunk muscles must have sufficient strength and endurance to satisfy the demands of spinal control, the efficacy of the muscle system is dependent on its controller, the central nervous system (eNS) (Panjabi 1992b). The challenge for the eNS to move and control the spine is immense, despite constant changes in internal and external forces. The eNS must continually interpret the status of stability, plan mechanisms to over­ come predictable challenges and rapidly initiate activity in response to unexpected challenges. It must interpret the afferent input from the peripheral mechanoreceptors, and other sensory systems, compare these requirements against an 'internal model of body dynamics' and then generate a coordinated response of the trunk muscles so that the mus­ cle activity occurs at the right time, at the right amount, and so on. To further complicate this issue, muscle activity must be coordinated to maintain control of the spine within a hierarchy of interdependent levels: control of intervertebral translation and rotation, control of spinal posture / orienta­ tion, control of body with respect to the environment. Finally, unlike the muscles of the limb, trunk muscles per­ form a variety of homeostatic functions in addition to movement and control of the trunk, including respiration and continence. This chapter reviews the elements that con­ tribute to the control and movement of the trunk, the strate­ gies used by the eNS to undertake this control and factors that complicate or compromise this control owing to con­ flict between trunk muscle functions and pain.

Optimal trunk function is a complex interplay between movement and control of the integrity of the spine and pelvis at the intersegmental level, at a global level involving


12 0


the control of orientation (e.g. control of lordosis, control of pelvic rotation), and the contribution of the trunk to mainte­ nance of equilibrium of the body with respect to gravity and other external forces (Fig. 10.1). All movements and pos­ tures are a complex interaction of movement and stability (Massion 1992). In reality, even static postures involve movement (for example small cyclical movements of the trunk and lower limbs compensate for disturbance to pos­ ture from respiration (Gurfinkel et al 1971, Hodges et al 2002a)), and movement occurs in conjunction with a subtle background of postural adjustments. Movement perturbs stability as a result of the interaction between internal and external forces (Massion 1992). These forces include the reactive moments from limb movements, changes in the influence of gravity on the body as a result of the modifica­ tion of the position of the centre of mass with movement and the interaction with objects and the environment (for example catching a ball). Even a simple action such as a movement of a limb changes the position of the centre of mass and is associated with reactive moments that are equal in amplitude but opposite in direction to the forces produc­ ing the moment. There is considerable argument about which parts of a task are movement related and which are purely posture related. In fact movement is used by the CNS to maintain stability and minimize energy expenditure. Rather than making the spine rigid, the CNS uses coordi­ nated movement to oppose and dissipate forces acting on the trunk. For instance, small movements of the trunk are initiated prior to limb movements that are opposed to the direction of reactive forces (Hodges et al 1999, 2000a), and rotation of the pelvis occurs around each orthogonal axis during gait (Perry 1992). Thus the control of movement and stability of the spine is complex. Moreover, the strategies

used by the CNS and the muscles involved vary between the three levels of control (intersegmental control, orienta­ tion control and control of body equilibrium). However, the understanding of the demands of stability is complicated by disagreement regarding the definition of the term 'stability'.
Models of stability

The most common contemporary view of spinal stability is based on the Euler model which considers the control of buckling forces (see, for example, Crisco & Panjabi 1991, Gardner-Morse et al 1995, Cholewicki & McGill 1996). This is based on the understanding that buckling failure of the lumbar spine, devoid of muscle, occurs with compressive loading of as little as 90 N (Lucas & Bresler 1960). This model argues that activity and stiffness of antagonistic muscles is required to maintain the lumbar spine in a mechanically stable equilibrium (Crisco & Panjabi 1991, Gardner-Morse et al 1995, Cholewicki & McGill 1996). Due to the emphasis on buckling, this element relates particu­ larly to the control of orientation and it has been argued that muscles act like guy wires to stiffen the intervertebral joints that they span (Crisco & Panjabi 1991). This definition is relatively static and suggests the maintenance of a set position of the spine. Few studies have considered this model in more dynamic terms (Cholewicki et aI1997). While control of buckling is a critical element of stability, there are additional factors to consider. Firstly, in terms of spinal health, this should be broadened to include the con­ trol of spinal movement; it is important to consider the control of the progression of changes in curvature and intervertebral motion. Secondly, the definition must incor­ porate control of the other components of stability, namely



Figur e 10.1

Multiple levels of tru nk control. A: Control of equilibriu m of the body. B: Control of trunk orientation. C: Intersegmental control.

Motor control of the trunk

12 1

the fine-tuning of intersegmental motion and the contribu­ tion of the trunk to postural equilibrium. Control of intersegmental translation and rotation is important, but cannot be completely separated from the control of spinal orientation and buckling forces (Panjabi et aI1989). Buckling can occur at the intervertebral level, but separate attention must be paid to control of translations and rotations. For instance, during an arc of movement it is important to control the coordination between translation and rotation at the intervertebral levels (Bogduk et al 1995). It has been shown that if stability of the spine is modelled with muscles of varying lengths, but leaving one segment with no muscle attachment, the spine remains unstable with stability equivalent to that achieved with no muscle at all, thus highlighting the importance of segmental attachment of the spinal muscles (Crisco & Panjabi 1991); segmental control is an essential component for spinal stability. At a more general level, as the trunk forms a large propor­ tion of the mass of the body, trunk movement is important for the control of postural equilibrium with respect to external forces. If the equilibrium of the body is disturbed by external forces (such as an unexpected movement of the support sur­ face) or internal forces (for example due to reactive forces from limb movement), movement of the trunk occurs to move the centre of mass over the base of support or alter the orientation of the body (see, for example, Horak & Nashner 1986, Keshner et al 1988). This stability function of the trunk is important to consider as it may influence the accuracy of control of spinal orientation or intervertebral motion. In par­ ticular, situations are likely to arise in which the requirement to move the trunk to restore balance may conflict with the demand to control the orientation of the spine. The same principles of control of orientation and inter­ segmental motion also apply to the pelvis. At one level there is the need to control orientation of the pelvis around the three orthogonal axes; however, there is also the requirement to control the relationship between segments of the pelvis. In upright positions the sacroiliac joint (SIJ) is subjected to considerable shear force as the mass of the upper body must be transferred to the lower limbs via the ilia (Snijders et al 1993, 1995). The body has two mecha­ nisms to overcome this shear force: one is dependent on the shape of the sacroiliac joint (form closure) and the frictional characteristics of the joint surface; the other mechanism involves generation of compressive forces across the SIJ via muscle contraction (force closure) (Snijders et al 1993, 1995). As with the spine, different muscles and recruitment strate­ gies are likely to be involved in control of each aspect of sta­ bility of the pelvis.
Control in Neutral

the trunk muscles and it has been argued that the region may increase (and thus the requirement for muscle activity) in situations of clinical instability (Panjabi 1992a).

Motor control of spinal stability requires an integrated sys­ tem that has sensors to detect the status of the body, a con­ trol system to interpret the requirements for stability and plan appropriate responses, and the muscles to execute the response. Consideration of these elements, in particular the architectural properties of the trunk muscles, is critical to understanding the mechanisms used by the nervous sys­ tem to control trunk muscles to coordinate movement and stability of the trunk.

The spine exhibits least stiffness around the neutral posi­ tion (Panjabi 1992a). Panjabi described this region of low stiffness as the 'neutral zone'. This region is important to consider as its stability is dependent on the contribution of

A large number of muscles have a mechanical affect on the spine and pelvis and all muscles are required to maintain optimal control. An important consideration is the redun­ dancy in the muscle system (i.e. many muscles cross the joints and may be capable of performing similar functions). However, there is considerable variation in the architectural properties of the trunk muscles, which has led to the pro­ posal by several authors that there may be functional dif­ ferentiation in the muscle system. This has implications for the potential contribution of these muscles to control and movement of the spine. In a general sense it is clear that the mechanical advantage of muscles to move and control the trunk varies due to factors such as the length of the moment arm and proximity to the joint, muscle attachments and the length and orientation of the muscle fascicles. Thus it has been argued variously that muscles are biomechanically more suited to either motion or stability (see, for example, Goff 1972, Janda 1978, Bergmark 1989, Richardson et al 1999, Sahrman 2002). In addition, as mentioned in the pre­ vious section, there are several elements to stability and there is likely to be some differentiation of contribution of muscles within this component. In reality there is likely to be a spectrum with muscles at the extremes that are ideally suited to control of intervertebral motion or spinal orienta­ tion and torque production; others in the middle of the spectrum make some contribution to both. Although sim­ ple division of muscles into groups is likely to oversimplify the complex control of lumbopelvic motion and stability, it provides a useful definition to consider as it contributes to our understanding of why the CNS uses different strategies to control the different muscle groups. Bergmark (1989) presented a model for the trunk that considered differentiation in the contribution of muscle to stability. This model identified muscles as either 'local' or 'global', based on anatomical characteristics (Fig. 10.2). The local muscles are those that cross one/few segments and have a limited moment arm to move the joint, but an ideal anatomy to control intervertebral motion. Bergmark





tures of the muscles that are intrinsic to the spine and those that lie superficially are presented in the following sections.
Intrinsic lumbopelvic muscles

Figure 10.2 Local and global muscles of the tru nk. A: Local mus­ cles attach d irectly to the spine and control intervertebral motion. B: G lobal mu scles transcend the spine and control spinal orientation.

included muscles such as the lumbar multifidus in this group; however, other muscles that satisfy these criteria are transversus abdominis (TrA) (Fig. 10.2A), intertransversarii and interspinales. In contrast, the global muscles have attachments to the pelvis and thorax and thus transcend multiple segments. These muscles have a larger moment arm and, thus, a larger torque generating capacity, and are suited to the control of orientation and balancing external forces. Examples of the global muscles include rectus abdo­ minis, obliquus externus abdominis, obliquus internus abdominis and the thoracic erector spinae. Muscles such as the lateral fibres of quadratus lumborum and parts of psoas also meet these criteria. There is considerable overlap between these systems with some muscles sharing features of both, such as the lumbar portions of longissimus and iliocostalis, which have one attachment to the lumbar ver­ tebrae and share some features of the local system. Considering this model, it is clear that optimal function of both systems is required to maintain spinal function. The local system has only a limited ability to influence the con­ trol of orientation and, similarly, the global system has only a limited ability to control intervertebral motion. In fact, the contribution made by the global system to the control of intervertebral motion occurs as a result of compressive forces exerted by co-activation of antagonist global mus­ cles. While compression can assist in the control of shear and rotation forces, this is associated with a cost: firstly, global co-activation increases the compressive load on lum­ bar segments (Gardner-Morse & Stokes 1998) resulting in increased intradiscal pressure and loading through the pos­ terior elements; secondly, antagonist global muscle co­ activation results in a restriction of spinal motion or rigidity of the spine and, as mentioned above, movement is an important component of optimal spinal control. In contrast, local muscles allow controlled spinal motion and have the ability to control individual segments rather than providing a general compressive force across the spine. Specific fea-

Transversus abdominis (TrA) is a sheet-like muscle that attaches from the inguinal ligament, iliac crest, thoraco­ lumbar fascia and the lower six ribs (Urquhart et al 2001). The attachment to the spine is via the three layers of the thoracolumbar fascia. The posterior layer of the fascia attaches to the spinous processes, the middle layer to the transverse processes and the anterior layer runs over quad­ ratus lumborum (Williams et al 1989). The contribution of TrA to spinal control is complex. Its muscle fibres have a rel­ atively horizontal orientation and therefore it has minimal ability to move the spine. However, it may contribute to rotation (Hemborg 1983, Cresswell et alI992, Urquhart et al 2002). Its contribution to spinal control is likely to involve its role in modulation of intra-abdominal pressure (IAP) and tensioning the thoracolumbar fascia. TrA has been. shown to be the abdominal muscle most closely associated with the control of IAP (Cresswell et al 1992, 1994) and recent data confirm that spinal stiffness is increased by lAP (Hodges et al 200lb, 2001d). Fascial tension may directly restrict intervertebral motion or provide gentle segmental compression via the posterior layer of the thoracolumbar fascia (Gracovetsky et alI985). Recent porcine studies con­ firm that the combined effect of IAP and fascial tension is required for TrA to increase intervertebral stiffness and the mechanical effect of its contraction on the mid-lumbar regions is reduced if the fascial attachments are cut (Hodges et al 2002b). For sacroiliac support, TrA acts on the lever formed by the ilia to increase anterior compression of the SIJ (Snijders et al 1995); this has been confirmed in vivo (Richardson et a12002). Multifidus has five fascicles that arise from the spinous process and lamina of each lumbar vertebra and descend in a caudolateral direction (Macintosh & Bogduk 1986). The most superficial fibres of each fascicle cross up to five seg­ ments and attach caudally to the ilia and sacrum. In con­ trast, the deep fibres attach from the inferior border of a lamina and cross a minimum of two segments to attach on the mamillary process and facet joint capsule (Lewin et al 1962). The superficial fibres are distant from the centres of rotation of the lumbar vertebrae, have an extension moment arm and can control the lumbar lordosis (Macintosh & Bogduk 1986). In contrast, the deep fibres have a limited moment arm and have only a minor ability to extend the spine (Panjabi et al 1989). While many trunk muscles are suited architecturally to the control of spinal orientation, most have a limited ability to control interver­ tebral shear and torsion (Panjabi et al 1989, Bogduk 1997). The deep fibres of multifidus are ideally placed to control these motions. Multifidus can control intervertebral motion by generation of intervertebral compression (Wilke et al 1995). The proximity of deep multifidus to the centre of rotation results in compression with minimal extension

Motor control of the trunk

12 3

moment to be overcome by antagonistic muscle activity. In addition, multifidus may contribute to the control of inter­ vertebral motion by control of anterior rotation and trans­ lation of the vertebrae (Macintosh & Bogduk 1986), or via tensioning the thoracolumbar fascia as it expands on con­ traction (Gracovetsky et al 1977). Several studies have pro­ vided in vitro and in vivo evidence of the ability of multifidus to control intervertebral motion (Kaigle et al 1995, Wilke et al 1995). Other muscles that share features with the intrinsic mus­ cles are the interspinales, intertransversarii, posterior fibres of psoas, medial fibres of quadratus lumborum and the lumbar portions of longissimus and iliocostalis. The inter­ spinales and intertransversarii are small muscles that have a high density of muscle spindles (see below) and have been argued to have an important sensory rather than motor function (Nitz & Peck 1986b). The posterior fibres of psoas that attach to the transverse processes of the lumbar vertebrae have a minimal moment arm for spinal move­ ment and have been argued to provide primarily an inter­ segmental compressive force (Bogduk et al 1992), and may have a primary function in intersegmental stability (Gibbons 2001). However, this requires clarification with EMG studies of this portion of the muscle. The medial fibres of quadratus lumborum, along with the lumbar erec­ tor spinae, have one attachment to the transverse processes of the lumbar spine and thus have a segmental attachment such that these muscles may contribute to both elements of spinal control and have been implicated in spinal stability (McGill et al 1996). Of the other abdominal muscles, obliquus internus has an attachment to the thoracolumbar fascia in a small proportion of people, thus providing a seg­ mental attachment to the spine (Bogduk 1997). Anteriorly this muscle has fibres that are parallel to those of TrA and may contribute to the force closure of the SIJ (Snijders et al 1995). However, despite the similarities to TrA there are dis­ tinct differences in control of these two muscles.
Superficiallumbopelvic muscles

obliquus externus and long erector spinae in this role (McGill 2002). Several authors argue that muscles such as the gluteus maximus may also contribute to the general control of the spine and generation of segmental compres­ sion (Vleeming et al 1995).

The contribution of the superficial muscles to lumbopelvic movement and stability is generally predictable based on the moment arm and direction of force provided by the muscles; that is, flexors generate flexion torque and oppose extension. Thus, in standing, the extensor muscles may be active to overcome trunk flexion due to gravity. However, it has been generally considered that antagonist trunk mus­ cles are co-activated to stiffen the spine and prevent buck­ ling (Gardner-Morse & Stokes 1998, McGill 2002). Muscles that provide this control include the oblique abdominal muscles, rectus abdominis, lateral fibres of quadratus lum­ borum, thoracic portions of the longissimus and iliocostalis. Furthermore, a contribution may also be provided by the lumbar erector spinae, superficial fibres of multifidus, medial fibres of quadratus lumborum, anterior fibres of psoas and latissimus dorsi. Recent studies using a Euler model have highlighted the important contribution of the

Multiple sensors contribute to the sensation of movement and position of the spine and pelvis. These include free nerve endings and receptors in the muscles, ligaments, annulus fibrosus, joint capsules and skin, with contribu­ tions from other senses such as vision and the vestibular and auditory systems. Muscle spindles are the most com­ plex of the mechanoreceptors and consist of sensory and contractile components that lie in parallel with muscle fibres so that they are stretched with the muscle (Gandevia et al 1992). The sensory component has two main types of sensory endings, bag and chain fibres. These endings are sensitive to length and/ or velocity of lengthening. The con­ tractile component of the muscle spindle provides a mech­ anism for the CNS to control the sensitivity of the muscle spindle and to adapt the spindle to changes in muscle length. The contractile component of the muscle spindle is innervated by a special class of motor neurons, called gamma motoneurons. It is considered that alpha and gamma motoneurons are co-activated during muscle con­ traction. Many studies have confirmed that the input from muscle spindles is critical for the perception of movement (Gandevia & McCloskey 1976), yet stimulation of single muscle afferents does not result in conscious perception (Macefield et al 1990). Spinal muscles have varying densi­ ties of muscle spindles; notably, the deep segmental mus­ cles have a high density of muscle spindles (Nitz & Peck 1986b) which is consistent with the proposal that these muscles have a critical role in sensation of intervertebral motion. Golgi tendon organs are located in series with the mus­ cle fibres in the tendon. These receptors provide an inhibitory input to the alpha motoneurons and were origi­ nally proposed to contribute only to strong contractions to prevent damage to the muscles. However, each receptor is attached to a small population of muscle fibres and is sen­ sitive to small forces to provide discrete detection of tension in different parts of the muscle (Houk & Simon 1967). Thus, these receptors are likely to provide an important contribu­ tion to feedback during movement. Joint receptors are encapsulated receptors (Ruffini end­ ings and pacinian corpuscles) situated in the joint capsule. The contribution of these receptors to perception of move­ ment and movement control has often been considered to be limited (Gandevia & McCloskey 1976). While some receptors are activated at specific ranges of motion, the majority fire at the end of range when the joint capsule is stretched (Nade et al 1987). Other joint structures such as the ligaments also contain receptors which may contribute

12 4


to proprioception. Mechanoreceptors are also present in the annulus of the disc (Roberts et al 1995). Electrical and mechanical stimulation of the mechanoreceptors in disc and other ligamentous structures modulates activity of muscles of the spine, including the multifidus muscle (see, for example, Indahl et al 1995, Solomonow et al 1998) (Fig. 10.3). There are several types of tactile receptors distributed in the layers of the skin. These receptors include pacinican corpuscles, Meissner corpuscles, Merkel cells and Ruffini endings and provide important tactile information. While input from the cutaneous receptors is important for the per­ ception of movement of large (e.g. knee, Edin 2001) and small joints (e.g. hand, Collins et al 2000) and is critical for the coordination of grip force (see Johansson & Westling 1988), it is not known whether this input contributes to con­ trol of the spine. The vestibular apparatus involves the saccule and utri­ cle, which detect the position of the head with respect to gravity, and the semicircular canals, which provide infor·· mation of acceleration of the head around the three major axes. The major function of the vestibular apparatus is to provide information about movements of the head. Integration of vestibular information and proprioceptive

information from the neck and trunk allow the interpreta­ tion of the position of the body relative to gravity. Interestingly, it has been argued that data from the control of the trunk are consistent with the presence of a gravity receptor in the trunk, in the region of the kidney, although the neural substrate of this mechanism is unclear (Mittelstaedt 1996). The visual and auditory systems provide information regarding the interaction between the body and the envi­ ronment or objects (Schmidt & Lee 1999). As such, vision provides an important contribution to control of movement and, although hearing does not play a major role in move­ ment control, auditory information may provide useful feedback from environmental factors and issues such as success of performance Genison 1997), for instance for feed­ back of the accuracy of movements involved in tasks such as foot contact during running. Although input from all sensory elements may provide information of disturbances to spinal stability, it is also crit­ ical to consider that sensory input is also required to pro­ vide input regarding the instantaneous status of the body and the internal and external forces acting on it, as well as development of an 'internal model' of the body and its dynamics so that the effect of movements and forces can be

Disc stimulation




EMG electrode site

Facet stimulation

l2 L3 L4

EMG electrode site


Stimulated side Contralateral side

Figure 10.3 Muscle response to electrical stimulation of the intervertebral disc and facet joint. Electrical stimulation (A) of mechanoreceptors is associated with a short latency response of the multifidus muscles (E). Adapted from Indahl et al 1995.

M otor co ntrol of the tru n k

12 5

predicted (Gahery & Massion 1981, Gurfinkel 1994). Input from all sources, including vestibular and proprioceptive, is required for the development, upkeep and interpretation of this model.

It is beyond the scope of this chapter to provide a detailed description of the organization of the control system. However, several important issues require consideration. Firstly, trunk muscles receive inputs from various parts of the eNS including corticospinal inputs (Plassman & Gandevia 1989), which to some extent, unlike the limb mus­ cles, course the spinal cord bilaterally or send collaterals to both sides (Kuypers 1981, Mori et al 1995). However, it is generally considered that there is more significant control of the trunk muscles by the brain stern and spinal structures (Kuypers 1981), for example the vestibulospinal and reticu­ lospinal systems. This is consistent with the relatively small size of the representation on the motor and sensory homunculi. The following section will consider the mecha­ nisms of cuntrol of the trunk muscles from a behavioural perspective, that is, consideration of the organization of muscle recruitment rather than consideration of the specific neural structure and events involved in their production.

stiffness and act as the first line of defence against an unex­ pected perturbation Gohansson et al 1991). This latter con­ trol strategy includes components of both feedforward and feedback mediated control. In general, normal function involves a complex combination of these strategies. As mentioned above, there is considerable redundancy in the motor system and multiple strategies could be used by the eNS in any given situation. The following sections outline evidence which argues that the eNS draws on the architec­ tural properties of trunk muscles in a specific manner to concurrently meet the demands of movement and control of stability (i.e. control of intervertebral motion, orientation and body equilibrium).
Open-loop control of the trunk

The eNS has two primary strategies for the control of the movement and stability of the body, including the trunk: feedforward or 'open'-loop strategies for situations in which the outcome of a perturbation is predictable and the eNS can plan strategies in advance; and feedback or 'closed'-loop strategies in which responses are generated in reaction to sensory input (visual, vestibular, proprioceptive input, etc.) from unpredictable perturbations (Schmidt & Lee 1999) (Fig. 10.4). In addition, due to time taken to initi­ ate a response to sensory input, the eNS may also generate an underlying level of tonic activity to increase the muscle
Closed-loop control system Open-loop control system


Afferenl input



Interpretation! error detection



Motor planning


I t
Motor command


Muscle activity

Figu re 10.4

Open- and closed-loop control systems.

Open-loop control implies that all aspects of the movement performance are pre-planned by the eNS and the move­ ment occurs without modification by sensory feedback (Fig. 10.4). Movements that are likely to fit into this cate­ gory are predictable ballistic and repetitive movements and predictable challenges to spinal control such as voluntary limb movements. Basic evidence that this type of control exists comes from studies of humans and animals with deafferented limbs. In these cases, limb movement can occur that is almost indistinguishable from that of a limb with a full complement of sensory input except for fine con­ trolled movements of the fingers, which appear slightly clumsy (Taub & Berman 1968). To reconcile these observa­ tions, theories have been developed of mechanisms of gen­ eration of movement patterns. In animals the presence of central pattern generators (ePG) has been confirmed (Grillner 1981). Basically, a ePG is a collection of neurons that may control a repetitive function such as locomotion or respiration. These neuron groups can control the alternat­ ing contraction of muscles to perform the movement and while they can be modified by afferent feedback they can function independently of feedback. The existence of ePGs has not been confirmed in humans. Another organizational theory to explain the central control of movement is the concept of the motor programme. The motor programme theory involves a memory based mechanism whereby a generalized motor programme is stored as an abstract representation of a group of move­ ments that are retrieved when a movement is performed (Schmidt & Lee 1999). This theory argues that the eNS stores details of invariant features of a movement (for example order of events, relative timing, relative force). This information is accessed, with selected task duration and muscles, when the movement is performed. There are several problems to consider: for instance, a large amount of information would need to be stored to cover the full complement of movement possibilities and there are a large number of degrees of freedom. This issue was highlighted by Bernstein (1967), who argued that there are too many components that need to be controlled concurrently. For



even the simplest movements of the hand, motion of each joint between the fingertip and the floor requires consider­ ation. This is compounded when considering all of the muscles that are available to control each joint and the motor units within each muscle. As suggested by Bernstein, this is an enormous problem for the CNS in view of the resources required to individually control the large number of muscles and joints. A system is needed that can reduce processing demands, for instance by grouping degrees of freedom together. Another model of movement control, the dynamic pat­ tern theory (Kelso 1984), has been presented to reconcile some of these difficulties in movement control. The dynamic pattern theory argues that there is no central rep­ resentation of all components of the movement, but instead the organization of the muscle contractions and joint move­ ment is coordinated by environmental invariants and limb dynamics. Central to this theory is the idea that movements are attracted to steady-state behaviours and movements follow the principles of non-linear dynamics. In other words, if a particular variable is changed systematically the system may move between separate stable states. A famil­ iar example to illustrate this point is the transition from walking to running. In the dynamic pattern theory it is argued that at slower speeds the movements of the arm and legs are 'attracted' to a coordinated pattern that is walking, yet at faster speeds the pattern changes, in part for reasons

of efficiency. Thus, coordinated movement is self-organized according to the characteristics of limb behaviour and envi­ ronmental constraints. Currently the debate continues regarding these two theories. In reality movement may be coordinated by a hybrid of both possibilities. Lumbopelvic stability is controlled in a feedforward or open-loop manner when the perturbation to the trunk is predictable. For instance, activity of the trunk muscles occurs in advance of the muscle responsible for movement of the upper (Belenkii et al 1967, Bouisset & Zattara 1981, Aruin & Latash 1995, Hodges & Richardson 1997b) and lower limbs (Hodges & Richardson 1997a) and prior to loading when a mass is added to the trunk in a predictable manner (Cresswell et a11994) (Fig. 10.5). In this type of task the CNS predicts the effect that this movement will have on the body and plans a sequence of muscle activity to over­ come this perturbation. This prediction involves an 'inter­ nal system of body dynamics' which is an abstract construct built up over a lifetime of movement experience and holds information of the interaction between internal and exter­ nal forces (Gurfinkel 1994). Several possibilities could explain the organization of the movement and postural parts of the task. In general the postural activity could exist as a part of the motor command for movement or the pos­ tural part could be organized separately, but in parallel with the movement command. Several studies have inves­ tigated this question and are generally in support of the

Figure 10.5 Feedforward control of trunk stability. Rapid arm movement is associated with a sequence of trunk muscle activity that varies between direc­ tions of limb movement. Onsets of activity of deltoid and the trunk muscles are shown. The deep muscle, transversus abdominis, is controlled separately and does not vary with movement direction. Adapted from Hodges Et Richardson 1996. Key: TrA transversus abdominis, 01 obliquus internus abdominis, OE obliquus externus abdominis, RA rectus abdominis, ES erector spinae.
= = = = =



Onset deltoid


� �T'A
__ OE

50 ms

M otor co ntro l of the tru n k


parallel process model (Massion 1992). An important fea­ ture of this feedforward control of the spine is that it pro­ vides . insight into the differential strategies used by the CNS to control each of the elements of stability and how these may be integrated. Consistent with the architectural properties of the trunk muscles described above (pp. 121-123), the temporal and spatial parameters of activity of the superficial trunk muscles are linked to the direction of forces acting on the spine (i.e. superficial trunk muscle activity is earlier and larger in amplitude when their activ­ ity opposes the direction of reactive forces), and thus con­ sistent with the control of orientation of the spine (Aruin & Latash 1995, Hodges & Richardson 1997b, Hodges et al 1999). In association with limb movements, this activity has also been shown to be consistent with the control of the dis­ turbance to equilibrium and to move the COM (centre of mass) in a manner consistent with the maintenance of upright stance (Aruin & Latash 1995, Hodges et aI1999). In contrast, activity of the deep intrinsic muscles (both TrA and multifidus) is independent of the direction of reactive forces (Hodges & Richardson 1997b, Moseley et al 2000). This is consistent with the architectural properties of these muscles to provide a general increase in intervertebral con­ trol. Thus, the data suggest that the CNS uses feedforward non-direction-specific activity of the intrinsic muscles to control intervertebral motion and tuned direction-specific responses of the superficial muscles to control spinal orien­ tation (Hodges & Richardson 1997b). Recent data suggest that the CNS uses discrete strategies to control each factor. When the preparation for movement is manipulated or subjects perform an attention demanding task, the latency for limb movement and the postural activity of the superfi­ cial muscles is delayed but there is no change in the latency of the deep muscle response (TrA, Hodges & Richardson 1999; deep fibres of multifidus, Moseley et al 200la). This suggests that the deep muscle response is more rudimen­ tary and may be controlled by a more basic mechanism by the CNS. Importantly, these responses have been shown to be linked to the speed of limb movement (Hodges & Richardson 1997c) and the mass of the limb (Zattara & Bouisset 1986, Hodges & Richardson 1997a), suggesting that the CNS predicts the amplitude of the reactive forces and adjusts the feedforward responses accordingly. Repetitive limb movements may also provide an exam­ ple of open-loop control. However, as the movement is ongoing it is not possible to exclude the contribution of afferent input to the organization of the trunk muscle activ­ ity, and studies have argued that spinal mechanisms dependent on afferent feedback may be important for this control (Zedka & Prochazka 1997). Although the mecha­ nism for control of repetitive movement is not completely understood, there is evidence of differential activity of the deep and superficial muscles that is consistent with the dif­ ferent roles of these muscles. For instance, tonic activity of the intrinsic spinal muscles occurs in association with repetitive upper limb movement (TrA, Hodges & Gandevia

2000b; multifidus, Moseley et a12002), repetitive lower limb movement during gait (Saunders et al 2002) and repetitive trunk movement (Cresswell et aI1992). In contrast, superfi­ cial muscle activity occurs in a phasic manner linked to the direction of limb movement.
Closed-loop control of the trunk

In a closed-loop system the command to move may be gen­ erated in a similar manner to an open-loop system; how­ ever, the intended movement is compared against feedback regarding the status of the body and its relationship to the environment (see Fig. 10.4). If the feedback differs from the intended movement an error command is generated to cor­ rect the movement performance. In this way sensory feed­ back is used to mould and correct movement performance (Schmidt & Lee 1999). Clearly this type of control requires effective systems for detecting the state of the environment and the position and movements of the body segments. These sensors were out­ lined above (see section on sensors, pp. 123-125). Although the concept of closed-loop control may be considered in terms of higher information processing and consciousness, this system may operate at a variety of levels from simple monosynaptic reflexes to complex fine motor tasks involv­ ing coordinated finger movements. It is important to con­ sider these different levels of control. At the more basic end of the spectrum, closed-loop con­ trol may operate at the reflex level. This may include mono­ synaptic stretch reflexes, which involve stretch of a muscle spindle generating afferent impulse from the receptor region of the spindles that excite the alpha motoneurons in the same muscle, resulting in contraction. Short-latency reflexes have been identified in the paraspinal muscles when subjects catch an unexpected mass in their hands (Wilder et al 1996, Leinonen et al 2001, Moseley et a12001b) and responses have been recorded in paraspinal (Dirnitrijevic et al 1980) and abdominal muscles (Kondo et al 1986, Myriknas et a12000) in response to a mechanical tap to the muscle. These reflex responses activate the paraspinal muscles en masse with no differentiation between deep and superficial components (Moseley et al 2001b). Simple responses are inflexible and represent a basic mechanism for the motor system to correct an error, for example to resist an imposed stretch. However, there appears to be some integration. For instance, reflex changes may occur in other related muscles, including contralateral muscles (Beith & Harrison 2001), and activity of TrA occurs prior to that of the paraspinal muscles when the trunk is unexpectedly flexed by addition of a mass to the front of the trunk (Cresswell et al 1994). Furthermore, activity of TrA and the paraspinal muscles occurs at the same time as the trunk is perturbed when a mass is added to the upper limbs during arm movement (Hodges et aI2001c). This lat­ ter finding suggests that afferent input from distant seg­ ments may be involved in initiation of the trunk muscle

12 8


response. When the predictability of the perturbation is increased and higher centre input may influence the response, the paraspinal muscles are differentially active, with earlier activity of deep multifidus (Moseley et al 200lb) (Fig. 10.6). This also occurs when paraspinal muscle activity is reduced when load is removed from the trunk, by removal of a load from the upper limbs (Hodges et al 2002b). This unloading response is commonly argued to be due to removal of the support for muscle contraction from spindle afferent input (Angel et al 1965, Nitz & Peck 1986a). Other basic responses have been identified in response to electrical and/ or mechanical stimulation of afferents in the ligaments, annulus, facet joint capsule and SIJ in pigs (see Fig. 10.3), cats and humans (Indahl et al 1995, 1997, 1999; Solomonow et al 1998, 1999). In general, activity of multi­ fidus was initiated with short latency on both sides and over multiple spinal segments in response to the stimulus. The nature of the response was affected by the site of stim­ ulation on the annulus (Holm et a12000) and SIJ (Indahl et a11999), and could be modified by injection of analgesic or saline into the facet joint capsule. These reflexes provide a strategy for mechanical stimulation of the spinal structures to influence trunk muscle activity in a reflex manner. Alternatively the response may modulate descending drive to the muscles. More complex than simple stretch reflexes are the long­ loop reflexes that involve information processing at higher levels of the eNS, including transcortical mechanisms. These responses have a longer latency than the simple stretch reflex, are more flexible and can be modified voluntarily (Marsden et aI1977). Due to their flexibility these responses are thought to have a greater role in error correction. Another response group are the triggered responses (Schmidt & Lee 1999). These responses are faster than a voluntary reaction

time but involve a more complex and widespread response than is initiated via simple reflex mechanisms. For instance, when the support surface on which a person is standing is rapidly moved, a complex interplay of several body seg­ ments, including response of trunk muscles, is initiated in order to maintain the equilibrium of the body (Horak & Nashner 1986, Keshner & Allum 1990). Two main strategies have been identified that involve either ankle movement (ankle strategy) or hip movement (hip strategy), depending on the context and the support surface characteristics (Horak & Nashner 1986). Trunk movement, and thus activation of the superficial trunk muscles, is a critical component of these strategies, particularly the hip strategy. The most complex level of closed-loop control is the fine control of long duration tasks that require accuracy. In these tasks, the sensory information may be used consciously to provide feedback of performance and continually modulate movement performance. However, even during these con­ scious goal-directed tasks, sensory information may be used at a subconscious level to modulate muscle activity.
Control of muscle stiffness

A third type of control strategy is related to both feedback and feedforward control and involves modulation of the 'tone' in specific muscles to provide an underlying degree of stability to the joints. This activity increases the stiffness of muscles that surround the joints (Bergmark 1989, Gardner-Morse et aI1995). Muscle stiffness is the property of muscles to act as springs (i.e. the ratio of length change to force change) and has viscoelastic and activity related components. Muscle stiffness provides control of forces applied to a joint and contributes to control before even the shortest reflex response could be initiated Gohansson et al

Figure 10.6 Feed back med i ated response of the back muscles to load ing of the trunk. When a load i s d ropped i nto the bucket held in the hand s (A). acti vity of the d eep, superfi cial and l ateral compo­ nents of the multifidus (onset indi cated by arrows) occurs with short latency after the perturbation to the trunk. When the perturbation is expected , the d eep and superficial fibres of multifidus are con­ trolled differentially. Reproduced from Moseley et al 2003. Key: Deep MF d eep fibres of multifid u s, Sup MF superficial fibres of multifid us, Lat MF lateral fibres of multifidus, ES T7 erector spinae at T7
= = = =


B Perturbation


Lat MF


Biceps '''-A>-.-w.l'-f

50 ms

Motor con trol of t h e t r u n k

12 9

1991) and it has been argued that postural stability may be controlled by modulation of stiffness of the ankle muscles (Winter et al 1998). Similarly, stability of the trunk may be controlled by stiffness of the spinal muscles. Importantly, the activity related component of muscle stiffness is modu­ lated by feedback from spindle and ligament afferents Gohansson et al 1991). It is the stretch reflex and the control of the gamma motoneurons, which control the sensitivity of the sensory component of the muscle spindles, that control this system. In addition, the reflex activity of multifidus muscle in response to stimulation of mechanoreceptors in the lumbar disc and ligaments (Indahl et al 1995, 1997, 1999) and supraspinous ligament in humans (Solomonow et al 1998) may contribute to stiffness control.
Integrated control of stability and movement of the trun k

issue in terms of the models of motor control of the trunk muscles presented in the previous section.
Chang es in open-l oopcontrolmechanism s

It is important to consider that all the processes defined above may act concurrently and the outcome of feedfor­ ward processes may be moulded by later feedback medi­ ated processes. In general, feedforward and feedback mediated responses closely match the demands of the task and are scaled to the amplitude of the perturbing forces and the context of the perturbation. As such, muscle activity directed to the control of stability represents a finely tuned component of human movement.

The major factor that has implicated changes in the open­ loop control of movement is changes in feedforward strate­ gies. As mentioned above, these strategies are pre-planned by the nervous system and represent the pattern of muscle activity initiated by the CNS in advance of movement. Several studies have investigated the onset of muscle activ­ ity in association with rapid limb movements (Hodges & Richardson 1996, 1998). These studies investigated people with chronic recurrent low back pain (LBP) when their pain was in remission. The most consistent finding was delayed activity of TrA with arm and leg movements in all direc­ tions (Fig. 10.7). Thus, activity of TrA was absent in the period before movement. This is consistent with a compro­ mise in the control of intervertebral motion (see section on models of stability). Activity of the superficial abdominal muscles was delayed only with specific movements. A major finding was that the change in TrA activity could not be explained by inhibition of the response or delayed transmission in the CNS, as the delay was different for each movement direction (i.e. there was a change in strategy, not a greater delay for the message to be transmitted to the motoneuron). Further studies have challenged the coordi­ nation of these responses, by manipulation of preparation for movement. These data suggest that the responses are a result of inappropriate motor planning rather than changes in excitability or transmission of the command in the CNS (Hodges, 200la) (see Fig. 10.10).

The delicate balance of motor control of the trunk may be compromised by a number of factors including pain and conflict between the multiple functions of the trunk mus­ cles. These factors present challenges to the motor control of the trunk muscles and may impair the control and sta­ bility of the lumbopelvic region.
The effect of pain and injury on motor control

Flexion TrA






r-tD--i J







: f-W-I *

f-b--.1 � � �

� J

Many studies have investigated changes in trunk muscle activity with acute and chronic pain. While most have eval­ uated the strength and endurance of the trunk muscles, this has led to variable results. For instance, some show reduced strength and endurance (see, for example, Suzuki et aI1977), while others do not (see, for example, Thorstensson & Arvidson 1982). It has been suggested that these changes may be more related to inactivity than pain (Thorstensson & Arvidson 1982). Furthermore, the importance of changes in strength and endurance is unclear as maximum strength and endurance are infrequently required in function and these parameters indicate little of how the muscles are used. Alternatively, studies have evaluated the control of the trunk muscles. It has been argued that impaired control of the trunk muscles may lead to inadequate support for the spine and pelvis, leading to injury and pain (Panjabi 1992b, Cholewicki et al 1997). This section considers this


: 1-0--1 : �*


: I



1 00


.,* , a
Time (ms)

J J J JI-O-l
100 -100

I f-;R;!

1 00


Figure 10.7 Group data for subjects with recurren t l ow back pai n and m atched control subjects showi n g the onset of EMG acti vity of the trun k muscles relative to that of deltoid with m ovement of the arm i n three directions. Zero indicates the onset of deltoid EMG. The onset of TrA acti vity is delayed i n low back pain subjects with move­ ment in each d i rection thus fai lin g to prepare the spine for the per­ turbati on from li mb movement. Adapted from H od ges Ii Richardson 1996. Key: TrA transversus abdom i nis, 01 obliquus i n tern us abdo­ min is, OE obliquus extern us abd om i n is, RA rectus abdom i n is, ES erector spi n ae, N LBP non low back pai n, LBP low back pai n .
= = = = = =

1 30


Changes in cl osed-l oopcontrolm echanism s

Changes in all elements of the closed-loop control system have been reported. However, as closed-loop control incor­ porates a complex interaction between input and output, in most studies it is difficult to determine the exact component or components of the system that are responsible for the change in motor control. For instance, if the amplitude of activity of a muscle is increased during a movement task it is difficult to determine whether the change results from inaccurate feedback from the periphery, inaccurate inter­ pretation of normal feedback or inability to initiate an appropriate command. However, in specific instances the component can be identified. The basis of closed-loop control is accurate feedback from movement. One of the most common of the motor control deficits that have been identified in association with lumbopelvic pain and injury is sensory deficit. This has been identified in two major ways, first by measurement of the acuity or smallest perceptible stimulation, such as the smallest movement that can be accurately detected, and secondly, the ability to accurately copy a position or return to a position of a limb after it has been demonstrated with the same or opposite limb. Using these methods studies have identified decreased acuity to spinal motion in low back pain (Taimela et a11999) and impaired ability to accu­ rately reposition with low back pain (Gill & Callaghan 1998, Brumagne et al 2000). Due to the importance of sensory information to closed-loop control of movement, deficits such as these may lead to impaired movement control at a number of levels. For instance, impaired acuity may lead to delayed reflex responses as a result of increased time to reach the threshold for movement detection. More complex changes are also possible, such as impaired coordination
F i g u re

during voluntary movement due to inaccurate feedback from movement. This inaccurate feedback may lead to faulty error detection and correction. Another possibility is that inaccurate feedback may lead to development of a faulty 'internal model of body dynamics'. In this case the CNS may generate commands that are inaccurate for per­ formance of the required movement. An additional possi­ bility is that muscle spindle sensitivity may be altered by pain (see, for example, Pedersen et aI 1997). The mechanism for sensory feedback to change with injury and pain may be multifactorial. For instance, it may be due to injury to joint, muscle or cutaneous receptors. Alternatively it may be due to changes in interpretation of the afferent input such as the potential for afferent input to be misinterpreted as nociceptive in hyperalgesia. In addi­ tion, changes in muscle activity may affect sensory acuity. Muscle activity is known to augment acuity (Gandevia et al 1992); thus any change in activation may adversely affect movement sensation. Furthermore, many muscles, particu­ larly the deep muscles close to the joints, have extensive attachments to joint structures and contraction is likely to affect sensation. Finally, several studies have argued that sensory acuity may be reduced by fatigue (Carpenter et al 1998); thus decreased muscle endurance with injury or pain may lead to impaired sensory acuity. Changes in a variety of reflex responses have been iden­ tified in musculoskeletal pain syndromes. These changes include delayed onset of activity of the erector spinae to trunk loading (Magnusson et a11996) and delayed offset of activity of the oblique abdominal and thoracolumbar erec­ tor spinae muscles of the trunk in response to unloading in chronic low back pain (Radebold et al 2000) (Fig. 10.8). However, others have failed to find changes in reflex
LOW BACK PAIN PATIENTS 0.6 Extension 0.5 0.4 0.3 0.2 Extension

front (exte nsion) or back (flexion) of the tru n k a n d is suddenly removed the tru n k m u scles m u st reduce their activity to m a i n ta i n the u p right positi on of the tru n k. When peo p l e have low back pa i n the offset of the extern a l o b l i q u e a b do m i n a l a n d t h o racic erector s p i n a e m u scles is d e layed. Reprod u ced from Radebold et al m i n is, EO 10 LE
= =


When a mass attached to the
0.6 0.5



0.4 03 0.2 0.1

g fil

o b l i q u u s extensor a b d o m i n is,


Key: RA


rectus a bdolatis-









o b l i q u u s i n ter n u s a b d o m i n i s, LD




10 ON






10 ON


s i m u s dorsi , TE

thoracic erector s p i n ae,
0.6 0.5 0.4

l u m b a r erector spi n ae.


0.5 0.4 0.3 0.2



g fil

0.3 0.2 0.1 0



10 OFF










Motor control of t h e tru n k


responses of the erector spinae, elicited by a muscle tap, with experimentally induced pain (Zedka et aI1999). Changes in control of trunk muscle activity occur during ongoing functional movements (i.e. closed-loop control). For instance, reduced amplitude of activity of multifidus has been identified during functional tasks in people with low back pain (Lindgren et a11993, Sihvonen et aI1997). In contrast, there has been considerable debate in the litera­ ture regarding the presence of augmented activity of the paraspinal muscles. In general these studies have had vari­ able results with studies reporting increased (Wolf & Basmajian 1977, Arena et aI1989), decreased (Sihvonen et al 1997), asymmetrical (Cram & Steger 1983) and no change in activity (Collins et al 1982). A consistent finding has been sustained activity of the erector spinae muscles at the end of range of spinal flexion, a point at which the erector spinae muscles are normally inactive (the 'flexion-relax­ ation' phenomenon), in people with low back pain (Shirado et al 1995). This has been replicated by experimental pain (Zedka et al 1999) (Fig. 10.9) and has been shown to limit intervertebral motion (Kaigle et aI1998). During gait, peri­ ods of silence in the erector spinae are reduced activity between heel contacts during gait (Arendt-Nielsen et al 1996). Additional evidence of hyperactivity of the superfi­ cial trunk muscles comes from the study by Radebold and colleagues (2000) that indicates delayed reduction of EMG activity when a load is removed from the trunk. Numerous studies have investigated parameters of ongoing closed-loop control of posture in people with low back pain. These studies have identified impairments of
Trunk displacement

balance when standing on one (Luoto et a11998) or two legs (Byl & Sinnott 1991) or sitting (Radebold et al 2001). Furthermore, an increased risk of low back pain or recur­ rence of pain has been identified for people with poor per­ formance in a test of standing balance (Takala & Viikari-Juntura 2000). These changes indicate a general reduction of the accuracy of the postural control system in these patients. Other more complex elements of control have also been found to be altered in low back pain. For instance, people with low back pain have a slower reaction time (Luoto et al 1995), and slow reaction time has been associated with musculoskeletal injuries (including low back pain) in a variety of sports (Taimela & Kujala 1992). Few studies have investigated the motor control of mul­ tifidus in LBP. However, changes in multifidus have been reported that may be indirectly associated with changes in control. For example, studies report changes in muscle fibre composition (Rantanen et al 1993), increased fatigability (Roy et al 1989, Biederman et al 1991), and reduced cross­ sectional area of multifidus has been identified as little as 24 hours after the onset of acute, unilateral LBP (Hides et al 1994). Thus, data appear to indicate that the deep local muscles and the superficial global muscles are commonly affected in an opposite manner by the presence of pain. Hypothe­ tically, this may result in reduced efficiency of interverte­ bral control. As mentioned earlier, the superficial muscles are inefficient for providing control at the intervertebral level and can only do so at the cost of increased spinal load­ ing and co-activation. As a result, a degree of the output of
Trunk displacement

°V (v :!l °
� C> Q) o

i n d u ced experi menta l l y b y i njec­ tion of hyperto n i c sa l i n e the nor­ m a l re laxation of the p a raspi n a l m uscles at the end o f tru n k flexion (i.e. flexion re laxation) ( m i d d l e panel) i s l o s t a n d m u scle activity is m a i ntai ned a lt h o u g h the ra nge of motion i s identical. Key:

Figure 10.9

When back pain is





10 12


0 -::- "-- 6-8 ---:- -12 � 2- -: ::- 10 4 =-




to r spinae. Adapted from Zedka Prochazka A, K n i g h t B, G i l la rd D a n d Ga uthier


Left ES


Left ES

M (1999).







Right ES


Right ES






1 32


these muscles must be diverted to intervertebral control. Thls is likely to compromise the ability of these muscles to deal with the control of orientation. Thls follows the hypothesis of Cholewicki et al (1997) who suggested that excessive activity in the superficial muscles might be a· measurable compensation for poor passive or active segmental support.
M echanismof chang es in m otor control

An important consideration is whether changes in motor control occur as a result of the pain (Fig. 10.10) or whether incompetent motor control strategies lead to inefficient spinal control, and thus microtrauma, nociceptor stimula­ tion and pain as suggested by Janda (1978) and Farfan (1973). While neither possibility can be ruled out, injection of hypertonic saline into the lumbar longissimus muscle to produce transient pain induced changes in the feedforward responses of TrA that are similar to those identified in clin­ ical pain (Hodges et al 200la). Changes in global muscle activity differed between individuals. However, in all sub­ jects, activity of at least one superficial trunk muscle was increased. This variability of the superficial muscles' response to pain is consistent with clinical observations. In separate studies, loss of relaxation of the erector spinae muscles has been replicated during trunk flexion (Zedka et a11999) and gait (Arendt-Nielsen et a11996) by experimen­ tally induced pain. However, it is likely that the motor con­ trol changes may also precede LBP. Several authors have argued that poor control may lead to microtrauma and eventual injury (Farfan 1973, Panjabi 1992b, Cholewicki et al 1997). Several studies have pro­ vided preliminary support for this hypothesis. For exam­ ple, Janda (1978) identified that many people with chronic back pain also had minor neurological signs, and people
F i g u re

with slow reaction times have been shown to have an increased risk of injury (Taimela & Kujala 1992). The mechanism for pain and nociceptor stimulation to affect motor control is poorly understood (see Fig. 10.10). Pain could affect motor output at any level of the motor system including the cortex, the motoneurons, reflex path­ ways and areas 'upstream' of the motor cortex involved in motor planning. Studies have identified changes in motoneuron excitability (Matre et a11998), decreased corti­ cal excitability (Valeriani et al 1999) and changes in sensi­ tivity of muscle spindles (Pedersen et a11997) in association with pain. However, the available data suggest that the change in motor control identified in LBP may be due to a change in motor planning, and not simple inhibition or transmission delays (Hodges 200la). Consistent with this hypothesis, pain changes the activity of areas of the brain involved in motor planning (see Derbyshire et al 1997 for a review). W hile the exact mechanism is unknown, pain may have a direct affect on motor planning or may affect plan­ ning as a result of the attention-demanding nature of pain or stress associated with pain. In terms of attention, it has been argued that changes may arise due to an inability to ignore unnecessary information and the affect that this would have on limited attention resources (Luoto et al 1999). However, recent data indicate that the changes in control with rapid arm movements cannot be replicated by attention-demanding or stressful tasks (Moseley et al 200la). However, fear of pain can replicate at least some features of the change in motor control identified with clin­ ical and experimental pain (Moseley et al 200la). These changes in motor control may be at least partially explained by the 'pain-adaptation' model. Thls model hypothesizes that movement velocity and amplitude is reduced in the presence of pain (Lund et al 1991). In terms

motor contro l .


Mecha n ism for pa i n to effect

Pain/nociceptor i nhibition

i- - - - - - - - __ Fear - - - 1 - - --, stress ,


_, p''' ning

_ _ _ � Cortical _

_ _ " I� _ model of body
_ ,.--

_ _ -'-_



� - - - - Motor - - .. I_ centres _ .. _ I� - - ..


Delayed central



Inaccurate input

Reflex inhibition

Motoneuron inhibition


Inaccurate 'virtual body'



Altered proprioceptive input


Motor control of t h e t r u n k


of limb and jaw movements, this is associated with reduced agonist activity and increased antagonist activity (Svensson et a11995). In terms of the control of trunk stability, this model may suggest increased co-activation of the trunk muscles to increase trunk stiffness. This would be consis­ tent with the prediction of Panjabi (1992b). As outlined above, one response of the nervous system to pain is aug­ mented activity of the superficial global muscles. In a pain-adaptation model this would be interpreted as an attempt by the CNS to splint and restrict motion of a region of the spine to protect it from injury or reinjury. As a result, the deep muscle activity may be redundant and reduced but at the expense of fine-tuning of segmental control. This hypothesis requires further investigation. Alternatively, pain may not affect motor control directly, but indirectly via the influence of pain on proprioception. In chronic pain, non-nociceptor mechanoreceptors may contribute to excitation of second order nociceptor neurons (Siddall & Cousins 1995) and pain may alter propriocep­ tive feedback (Capra & Ro 2000). Thus, pain may affect motor planning indirectly via inaccurate feedback and may influence feedforward responses as a result of devel­ opment of an internal model of body dynamics that is built on faulty input. A final factor to consider is that motoneuron excitability may be altered in the presence of pain and injury. One fac­ tor that may change motoneuron excitability is reflex inhi­ bition. The mechanism for reflex inhibition is generally considered to involve inhibition of the alpha motoneuron as a result of afferent input from effusion (Stokes & Young 1984) or injury to joint structures (Ekholm et al 1960). For instance, when effusion is present in the knee the motoneu­ ron excitability of quadriceps muscles is reduced (Spencer et al 1984). Furthermore, this affects certain muscles to dif­ ferent degrees, such as the oblique fibres of vastus medialis being inhibited with lower volumes of effusion than other vasti muscles. Reflex inhibition has also been argued to explain the rapid atrophy of multifidus in people with acute low back pain (Hides et al 1994), although this requires clarification.
Task conflict of the trunk muscles


su rro u n d the abdom i n a l cavity a re coord i n a ted for control of l u mbopelvic sta b i l ity, respiration and conti n e n ce.

Figu re 10. 1 1

Abd o m i n a l ca n ister. Activity o f t h e m uscles that

during arm movements (Hodges et al 1997a, 2002d, Hodges & Gandevia 2000a, 2000b). However, their involvement in spinal control presents a challenge to the CNS to coordinate the respiratory and continence functions. To further com­ plicate this system, respiration also presents a cyclical chal­ lenge to stability of the trunk and body equilibrium (Gurfinkel et al 1971).
R esp iration

Unlike limb muscles, the muscles of the trunk are involved in functions other than control and movement, such as res­ piration, continence and control of the abdominal contents. This introduces a challenge to the control system to coordi­ nate these functions. As mentioned above (section on intrinsic lumbopelvic muscles), the contribution of TrA to lumbopelvic stability involves increased lAP and fascial tension. Changes in these parameters require co-activation of the diaphragm and pelvic floor muscles, which control displacement of the abdominal contents. Co-activation of these muscles has been termed the 'abdominal canister ' (Hodges 1999) (Fig. 10.11). Studies have confirmed that activity of these muscles occurs in conjunction with TrA

Normal quiet respiration involves cyclical activity of the diaphragm, parasternal intercostal and scalene muscles during inspiration, with expiration generated passively by the elastic recoil of the lung and chest wall (DeTroyer & Estenne 1988). However, when the demand for respiration is increased and the rate and depth of expiration are increased, abdominal muscles are phasically activated dur­ ing the expiratory phase (Campbell 1952). If respiration is increased involuntarily (as in hypercapnoea) TrA is recruited at lower minute ventilation than the other abdom­ inal muscles (DeTroyer et al 1990, Hodges et al 1997b). Recent data indicate that this may vary between regions of the abdominal wall, with activity of the mid-region of TrA recruited with lower respiratory demand (Urquhart and Hodges, unpublished observations). Recent studies of repetitive limb movements confirm that when the arm is

1 34


moved repetitively to challenge the stability of the spine, tonic activity of the diaphragm and TrA is sustained, but is modulated with respiration to meet respiratory demands (Hodges & Gandevia 2000a, 2000b). In a mechanical sense, the diaphragm and TrA co-contract tonically, yet during inspiration diaphragm activity is increased and shortens (concentric), and TrA decreases its activity and lengthens (eccentric). The converse pattern occurs during expiration. Recent data confirm that this coordination also occurs dur­ ing natural repetitive movements such as locomotion (Saunders et al 2002). This coordination occurs as if there is summation of the respiratory and postural drives to these muscles, which may occur at the motoneuron, providing a mechanism for the CNS to coordinate these functions. However, when respiratory drive is increased by respira­ tory disease (Hodges et al 2000b) or by breathing with an increased dead space to induce hypercapnoea (Hodges et al 2001e) this coordination is compromised and tonic activity of the diaphragm and TrA is reduced. Respiratory movements of the ribcage and abdomen also generate a cyclical disturbance to stability of the trunk and body equilibrium. However, most studies have failed to identify a cyclical disturbance to the centre of pressure at the ground with respiration (Gurfinkel et a11971, Bouisset & Duchene 1994). This is due to small amplitude cyclical movements of the lumbar spine, pelvis and lower limb that are time-locked to respiration that match and counteract the disturbance to postural stability (Gurfinkel et al 1971, Hodges et al 2002a). Importantly, this postural compensa­ tion does not occur when people have low back pain (Guillemot & Duplan 1995, Grimstone & Hodges 2003).
Contin enc e

well as performing coordinated movement of the trunk. Theoretically, this coordination may also compromise the accuracy of stability. For instance, when body equilibrium is disturbed, movement of the trunk is required to maintain the position of the centre of mass over the base of support, and this demand may be inconsistent with the demand to maintain stability. Although in specific situations the trunk muscle activity has been found to be consistent with both tasks (Hodges et a11999), this may not be the case in all sit­ uations. For instance, if the support surface is moved when a mass is being lifted, conflict between postural and move­ ment tasks may arise. In this situation postural control has been shown to be compromised (Oddsson et al 1999, Huang et al 2001).
Im lications of task conflict p

Task conflict has important clinical implications for low back pain patients. It has been argued that respiratory and genitourinary problems are common in people with low back pain (Hurwitz & Morgenstern 1999, Finkelstein 2002) and this may compromise the normal coordination of pos­ tural, respiratory and continence functions of the trunk muscles. Thus, normal control of lumbopelvic stability and movement may be challenged by potential conflict between the multiple functions of the trunk muscles. This may lead to compromised accuracy of control.
Additional control issues

Similar to the challenge to respiration, the CNS must deal with the challenge to coordinate continence and spinal sta­ bility. Importantly, when intra-abdominal pressure is increased in association with contraction of the abdominal muscles, activity of the pelvic floor muscles is required to maintain continence. Numerous studies have confirmed that pelvic floor muscle activity occurs in conjunction with coughing (Deindl et al 1993) and lifting (Hemborg et al 1985) and recent data confirm that activity of the pelvic floor muscles precedes single limb movements in a non­ direction-specific manner (similar to TrA and deep multi­ fidus) and are tonically active during repetitive movements of the arm (Hodges et al 2002c). Other studies argue that voluntary activity of the pelvic floor muscles is associated with involuntary recruitment of TrA (Sapsford et al 2001, Critchley 2002) and, conversely, TrA activity is associated with pelvic floor muscle recruitment (Sapsford & Hodges 2001).
Oth er factors l eading to task conflict

Several other factors present challenges to motor control of the trunk, namely the function of adjacent segments and the role of the trunk as a reference frame. Irrespective of the stability of the trunk, it has been argued from a largely clin­ ical perspective that stability cannot be maintained in func­ tion if the motion of the adjacent joints is compromised, such that lumbar motion must compensate for reduced hip or thoracic flexibility. There is some evidence of this in the literature. For instance, hip range of motion has been shown to be reduced in people with low back pain (Ellison et aI 1990). The second additional factor that complicates the control of the trunk is that the CNS may use the trunk as a 'refer­ ence frame'. That is, the CNS may interpret the position of other regions with respect to the trunk. For . instance, dancers have been shown to control the lower limb in rela­ tion to the trunk (Mouchnino et al 1990, 1993). If true, opti­ mal control of the trunk has implications for coordination of regions other than the trunk. This requires further inves­ tigation.

As mentioned above (section on models of stability), the trunk muscles contribute to control of intervertebral motion, trunk orientation and whole-body equilibrium as

In summary, multiple strategies are used by the CNS to coordinate movement and control of the lumbopelvic region. A major issue is the numerous factors that can lead to compromise of the efficiency of the control system,

---,--- - ----

Motor control of t h e t r u n k


particularly of the deep local muscles of the region. It is activity of the deep muscles that is most commonly found to be ·impaired in the presence of pain and by conflict with other concurrent homeostatic functions. Although the deep muscles are not sufficient to provide control to the lumbar spine and pelvis, they provide a critical contribution, along with the superficial global muscles. Hypothetically, aug­ mented activity of the superficial muscles (at the expense of the deep muscles) may compromise the quality of spinal control as these muscles have a limited ability to fine-tune intervertebral motion and their activity is associated with the cost of reduced flexibility of spinal motion due to co­ contraction to counteract the torque output of these mus­ cles. Furthermore, it may be argued that dependence on the superficial muscles may compromise other functions such

a s respiration due to the attachments to the thorax and ribcage. In contrast, normal control of the deep local mus­ cles is likely to provide an efficient mechanism to control intervertebral motion without restricting spinal movement and without compromise to respiration. Thus, techniques to rehabilitate the coordination between these systems and motor control strategies can be justified.

KEYWORDS stability open loop closed loop sensor controller task conflict

Angel R W, Eppler W, Iannone A 1965 Silent period produced by unloading of muscle during voluntary contraction. Journal of Arena J G, Sherman R A, Bruno G M, Young T R 1989 Physiology 180: 864-870 Campbell E J M 1952 An electromyographic study of the role of the abdominal muscles in breathing. Journal of Physiology (London) 117: 222-233 Capra N F, Ro J Y 2000 Experimental muscle pain produces central spindles. Pain 86: 151-162 modulation of proprioceptive signals ariSing from jaw muscle

Electromyographic recordings of 5 types of low back pain subjects and non-pain controls in different positions. Pain 37: 57-65

Arendt-Nielsen L, Graven-Nielsen T, Svarrer H, Svensson P 1996 The during gait: a clinical and experimental study. Pain 64: 231-240 in feed-forward postural reactions during fast voluntary arm Beith I 0, Harrison P J 2001 Reflex control of the human internal Belenkii V, Gurfinkel V S, Paltsev Y 1967 Elements of control of voluntary movements. Biofizika 12: 135-141 movements. Experimental Brain Research 103: 323-332 influence of low back pain on muscle activity and coordination

Carpenter J E, Blasier R B, Pellizzon G G 1998 The effects of muscle fatigue on shoulder joint position sense. American Journal of Sports Cholewicki J, McGill S M 1996 Mechanical stability of the in vivo lumbar spine: implications for injury and chronic low back pain. Cholewicki J, Panjabi M M, Khachatryan A 1997 Stabilizing function of trunk flexor-extensor muscles around a neutral spine posture. Spine Collins G A , Cohen M J , Naliboff B 0 , Schandler S L 1982 Comparative analysis of paraspinal and frontalis EMG, heart rate and skin conductance in chronic low back pain patients and normals to various postures and stresses. Scandinavian Journal of Collins 0 F, Refshauge K M, Gandevia S C 2000 Sensory integration in joint. Journal of Physiology 529(2): 505-515 the perception of movements at the human metacarpophalangeal Rehabilitation Medicine 14: 39-46 22: 2207-221 2 Clinical Biomechanics 1 1 : 1-15 Medicine 26: 262-265

Aruin A S, Latash M L 1995 Directional specificity of postural muscles

oblique muscles. Abstracts - Society for Neuroscience 27: 930-933

Bergmark A 1989 Stability of the lumbar spine: a study in mechanical Bernstein N 1967 The co-ordination and regulation of movements. Biederman H J, Shanks G L, Forrest W J, Inglis J 1991 Power spectrum analysis of electromyographic activity: discriminators in the differential assessment of patients with chronic low back pain. Spine Bogduk N 1997 Clinical anatomy of the lumbar spine and sacnun. Bogduk N, Pearcy M, Hadfeild G 1992 Anatomy and biomechanics of Bogduk N, Amevo B, Pearcy M 1 995 A biological basis for instantaneous centres of rotation of the vertebral column. Proceedings of the Institution of Mechanical Engineers Part H : 1 77-183 Bouisset S, Duchene J L 1994 Is body balance more perturbed by 957-960 psoas major. Clinical Biomechanics 7: 109-119 Churchill Livingstone, London 16: 1179-1184 Pergamon Press, Oxford engineering. Acta Orthopaedica Scandinavica 60: 1-54

Cram J R, Steger J C 1 983 EMG scanning in the diagnosis of chronic Cresswell A G, Grundstrom H, Thorstensson A 1992 Observations on intra-abdominal pressure and patterns of abdominal intra-muscular Cresswell A G, Oddsson L, Thorstensson A 1994 The influence of sudden perturbations on trunk muscle activity and intra­ 98: 336-341 abdominal pressure while standing. Experimental Brain Research activity in man. Acta Physiologica Scandinavica 144: 409-418 pain. Biofeedback and Self-Regulation 8: 229-241

respiration in seating than in standing posture? Neuroreport 5:

Crisco J J, Panjabi M M 1991 The intersegmental and multisegmental muscles of the lumbar spine: a biomechanical model comparing Critchley 0 2002 Instructing pelvic floor contraction facilitates transversus abdominis thickness increase during low-abdominal Deindl F, Vodusek 0, Hesse U, Schussler B 1993 Activity patterns of Journal of Urology 72: 46-51 pubococcygeal muscles in nulliparous continent women. British hollowing. Physiotherapy Research International 7: 65-75 lateral stabilising potential. Spine 7: 793-799

Bouisset S, Zattara M 1981 A sequence of postural adjustments Brumagne S, Cordo P, Lysens R, Verschueren S, Swinnen S 2000 The role of paraspinal muscle spindles in lumbosacral position sense in individuals with and without low back pain. Spine 25: 989-994 Byl N N, Sinnott P L 1991 Variations in balance and body sway in middle-aged adults: subjects with healthy backs compared with subjects with low back dysfunction. Spine 16: 325-330 precedes voluntary movement. Neuroscience Letters 22: 263-270

Derbyshire S W, Jones A K, Gyulai F, Clark S, Townsend 0, Firestone L L produces differential patterns of central activity. Pain 73: 431-445 1997 Pain processing during three levels of noxious stimulation

1 36


DeTroyer A, Estenne M 1988 Functional anatomy of the respiratory muscles. In: Belman M (ed) Respiratory muscles: Function in health and disease. W B Saunders, Philadelphia, Vol 9, pp 1 75-195 DeTroyer A, Estenne M, Ninane V, VanGansbeke D, Gorini M 1990 Transversus abdominis muscle function in humans. Journal of Applied Physiology 68: 1010-1016 Dimitrijevic M R, Gregoric M R, Sherwood A M, Spencer W A 1980 Reflex responses of paraspinal muscles to tapping. Journal of Neurology, Neurosurgery, and Psychiatry 43: 1112-1118 Edin B 2001 Cutaneous afferents provide information about knee joint movements in humans. Journal of Physiology 531: 289-297 Ekholm J, Eklund G, Skoglund S 1960 On reflex effects from knee joint of cats. Acta Physiologica Scandinavica 50: 167-174 Ellison J B, Rose S J, Sahrmann S A 1990 Patterns of hip rotation range of motion: a comparison between healthy subjects and patients with low back pain. Physical Therapy 70: 537-541 Farfan H F 1973 Mechanical disorders of the low back. Lea and Febiger, Philadelphia Finkelstein M 2002 Medical conditions, medications, and urinary incontinence: analysis of a population-based survey. Canadian Family Physician 48: 96--1 01 Gahery Y, Massion J 1981 Co-ordination between posture and movement. Trends in Neuroscience 4: 199-202 Gandevia S C, McCloskey D I 1976 Joint sense, muscle sense, and their combination as position sense, measured at the dorsal interphalangeal joint of the middle finger. Journal of PhYSiology 260: 387-407 Gandevia S C, McCloskey D I, Burke D 1992 Kinaesthetic signals and muscle contraction. Trends in Neurosciences 15: 62-65 Gardner-Morse M G, Stokes I A 1998 The effects of abdominal muscle coactivation on lumbar spine stability. Spine 23: 86--9 1 Gardner-Morse M, Stokes I A F, Laible J P 1995 Role of muscles in lumbar spine stability in maximum extension efforts. Journal of Orthopaedic Research 13: 802-808 Gibbons S 2001 Biomechanics and stability mechanisms of psoas major. In: Vleeming A, Mooney V, Gracovetsky S A et al (eds) Fourth Interdisciplinary World Congress on Low Back Pain Pelvic Pain. European Conference Organisers, Montreal, Canada Gill K P Callaghan M J 1998 The measurement of lumbar , proprioception in individuals with and without low back pain. Spine 23: 371-377 Goff B 1972 The application of recent advances in neurophysiology to Miss M. Rood's concept of neuromuscular facilitation. PhYSiotherapy 58: 409-415 Gracovetsky S, Farfan H F, Lamy C 1977 A mathematical model of the lumbar spine using an optimised system to control muscles and ligaments. Orthopedic Clinics of North America 8: 135-153 Gracovetsky S, Farfan H, Helleur C 1985 The abdominal mechanism. Spine 10: 317-324 Grillner S 1981 Control of locomotion in bipeds, tetrapods, and fish. In: Brookhart M, Mountcastle V B (eds) Handbook of phYSiology. The nervous system. Motor control. American Physiological Society, Washington D C, Vol 2, pt 2, pp 1 179-1235 Grimstone S K, Hodges P W 2003 Impaired postural compensation for respiration in people with recurrent low back pain. Experimental Brain Research 151: 218-224 Guillemot A, Duplan B 1995 Etude de la prevalence des troubles posturaux au sein d'une cohorte de 106 1umbalgiquies. In: Gagey P Gurfinkel V S 1994 The mechanisms of postural regulation in man. Soviet Scientific Reviews. Section F. Physiology and General Biology 7: 59--89 Gurfinkel V S, Kots Y M, Paltsev E I, Feldman A G 1971 The compensation of respiratory disturbances of erect posture of man as an example of the organisation of interarticular interaction. In: Gelfard I M, Gurfinkel V S, Formin S V, Tsetlin M L, (eds) Models of the structural flmctional organisation of certain biological systems. MIT Press, Cambridge, Massachusetts, pp 382-395 W B (ed) Entrees du systeme postural fin. Masson, Paris, pp 71-77

Hemborg B 1983 Intraabdominal pressure and trunk muscle activity during lifting. Department of Physical Therapy. University of Lund, Lund Hemborg B, Moritz U, Lowing H 1985 Intra-abdominal pressure and trunk muscle activity during lifting. IV: The causal factors of the intra-abdominal pressure rise. Scandinavian Journal of Rehabilitation Medicine 17: 25-38 Hides J A, Stokes M J, Saide M, Jull G A, Cooper D H 1994 Evidence of lumbar multifidus muscle wasting ipsilateral to symptoms in patients with acute/subacute low back pain. Spine 19: 165-177 Hodges P W 1999 Is there a role for transversus abdominis in lumbo­ pelvic stability? Manual Therapy 4: 74--86 Hodges P W 2001a Changes in motor planning of feedforward postural responses of the trunk muscles in low back pain. Experimental Brain Research 141: 261-266 Hodges P, Gandevia S 2000a Activation of the human diaphragm during a repetitive postural task. Journal of Physiology 522: 165-175 Hodges P, Gandevia S 2000b Changes in intra-abdominal pressure during postural and respiratory activation of the human diaphragm. Journal of Applied Physiology 89: 967-976 Hodges P W, Richardson C A 1996 Inefficient muscular stabilisation of the lumbar spine associated with low back pain: a motor control evaluation of transversus abdominis. Spine 21: 2640--2650 Hodges P W, Richardson C A 1997a Contraction of the abdominal muscles associated with movement of the lower limb. Physical Therapy 77: 132-144 Hodges P W, Richardson C A 1997b Feedforward contraction of transversus abdominis is not affected by the direction of arm movement. Experimental Brain Research 114: 362-370 Hodges P W, Richardson C A 1997c Relationship between limb movement speed and associated contraction of the trunk muscles. ErgonOmics 40: 1220-1230 Hodges P W, Richardson C A 1998 Delayed postural contraction of transversus abdominis associated with movement of the lower limb in people with low back pain. Journal of Spinal Disorders 11: 46--56 Hodges P W, Richardson C A 1999 Transversus abdominis and the superfiCial abdominal muscles are controlled independently in a postural task. Neuroscience Letters 265: 91-94 Hodges P W, Butler J E, McKenzie D, Gandevia S C 1997a Contraction of the human diaphragm during postural adjustments. Journal of Physiology 505: 239-248 Hodges P W, Gandevia S C, Richardson C A 1997b Contractions of specific abdominal muscles in postural tasks are affected by respiratory maneuvers. Journal of Applied Physiology 83: 753--760 Hodges P W, Cresswell A G, Thorstensson A 1999 Preparatory trunk motion accompanies rapid upper limb movement. Experimental Brain Research 124: 69-79 Hodges P W, Cresswell A G, Daggfeldt K, Thorstensson A 2000a Three dimensional preparatory trunk motion precedes asymmetrical upper limb movement. Gait and Posture 11: 92-101 Hodges P W, McKenzie D K, Heijnen I, Gandevia S C 2000b Reduced contribution of the diaphragm to postural control in patients with severe chronic airflow limitation. Proceedings of the Annual Scientific Meeting of the Thoracic Society of Australia and New Zealand, Melbourne, Australia Hodges P, Moseley G, Gabrielsson A, Gandevia S 2001a Acute experimental pain changes postural recruitment of the trunk muscles in pain-free humans. Abstracts - Society for Neuroscience 27: 304--3 11 Hodges P W, Cresswell A G, Daggfeldt K, Thorstensson A 2001b In vivo measurement of the effect of intra-abdominal pressure on the human spine. Journal of Biomechanics 34: 347-353 Hodges P W, Cresswell A G, Thorstensson A 2001c Perturbed arm movements cause short latency postural responses in trunk muscles. Experimental Brain Research 138: 243--250 Hodges P W, Eriksson A E M, Shirley D, Gandevia S C 2001d Lumbar spine stiffness is increased by elevation of intra-abdominal pressure.

Motor control of t h e tru n k

1 37

Proceedings of the International Society for Biomechanics, Zurich, Hodges P W, Heijnen I, Gandevia S C 2001e Reduced postural activity Journal of Physiology 537: 999-1008 of the diaphragm in humans when respiratory demand is increased. Switzerland

Multiple muscle systems: biomechanics and movement organization. Springer-Verlag, New York, pp 481-497 Keshner E A, Woollacott M H, Debu B 1988 Neck, trunk and limb muscle responses during postural perturbations in humans. Experimental Brain Research 7 1 : 455-466

Hodges P, Gurfinkel V S, Brumagne S, Smith T, Cordo P 2002a Coexistence of stability and mobility in postural control: evidence from postural compensation for respiration. Experimental Brain Research 144: 293-302

Kondo T, Bishop B, Shaw C F 1986 Phasic stretch reflex of the abdominal muscles. Experimental Neurology 94: 1 20-140 Kuypers H 1981 Anatomy of the descending pathways. In: Brookhart J, Motor control. American Physiological SOCiety, Bethesda, MD, vol 2, Mountcastle V (eds) Handbook of Physiology. The nervous system.

Hodges P Moseley G, Gandevia S 2002b Differential control of the deep , from higher centres. Seventh International PhYSiotherapy Congress, Hodges P W, Sapsford R R, Pengel H M 2002c Feedforward activity of the pelvic floor muscles precedes rapid upper limb movements. Seventh International PhYSiotherapy Congress, Sydney, Australia Thorstensson A 2003 Intervertebral stiffness of the spine is increased in vivo porcine studies. Spine 28: 2594-2601 Sydney . and superficial compartments of multifidus is dependent on input

pt 1, pp 597-666

Leinonen V, Kankaanpaa M, Luukkonen M, Hanninen 0, Airaksinen 0, Taimela S 2001 Disc herniation-related back pain impairs feed­ forward control of paraspinal muscles. Spine 26: E367-372

Lewin T, Moffett B, Viidik A 1962 The morphology of the lumbar synovial Lindgren K-A, Sihvonen T, Leino E, Pitkanen M, Manninen H 1993 joints. Acta Morphologica Neerlanco Scandinavica 4: 299-319

Hodges P Kaigle-Holm A, Holm S, Erstrom L, Cresswell A, Hansson T, , by evoked contraction of transversus abdominis and the diaphragm: Holm S, Indahl A, Kaigle A, Gronblad M, Hansson T 2000 The

Exercise therapy effects on functional radiographic findings and segmental electromyographic activity in lumbar spine instability. Archives of Physical Medicine and Rehabilitation 74: 933-939

neuromuscular role of mechanoreceptors in the porcine lumbar Study of the Lumbar Spine, Adelaide, Australia, p 263

Lucas D B, Bresler B 1960 Stability of the ligamentous spine. Technical California, Berkeley and San Francisco Report esr. 11, no. 40. Biomechanics Laboratory, University of

intervertebral disc. Proceedings of the International Society for the

Horak F, Nashner L M 1986 Central programming of postural Journal of Neurophysiology 55: 1369-1381

Lund J P, Donga R, Widmer C G, Stohler C S 1991 The pain-adaptation model: a discussion of the relationship between chronic musculoskeletal pain and motor activity. Canadian Journal of

movements: adaptation to altered support-surface configurations.

Physiology and Pharmacology 69: 683-694

Houk J, Simon W 1967 Responses of Golgi tendon organs to forces applied to muscle tendon. Journal of Neurophysiology 30: 1466-1481

Luoto S, Heliovaara M, Hurri H, Alaranta H 1995 Static back endurance Luoto S, Aalto H, Taimela S, Hurri H, Pyykko I, Alaranta H 1998 One­ footed and externally disturbed two-footed postural control in a controlled study with follow-up. Spine 23: 2081-2089 patients with chronic low back pain and healthy control subjects: and the risk of low-back pain. Clinical Biomechanics 10: 323-324

Huang Q M, Hodges P W, Thorstensson A 2001 Postural control of the trunk in response to lateral support surface translations during 552-559 trunk movement and loading. Experimental Brain Research 141:

Hurwitz E, Morgenstern H 1999 Cross-sectional associations of asthma, hay fever, and other allergies with major depression and low-back pain among adults aged 20-39 years in the United States. American

Luoto S, Taimela S, Hurri H, Alaranta H 1 999 Mechanisms explaining processing: a controlled study with follow-up. Spine 24: 255-261

the association between low back trouble and deficits in information

Journal of Epidemiology 150: 1107-1116

Macefield G, Gandevia S C, Burke D 1990 Perceptual responses to

Indahl A, Kaigle A, Reikeras 0, Holm S 1995 Electromyographic response of the porcine multifidus musculature after nerve stimulation. Spine 20: 2652-2658 Indahl A, Kaigle A M, Reikeras 0, Holm S H 1997 Interaction between the porcine lumbar intervertebral disc, zygapophysial joints, and paraspinal muscles. Spine 22: 2834-2840

microstimulation of single afferents innervating joints, muscles and skin of the human hand. The Journal of Physiology 429: 113-129

McGill S 2002 Low back disorders: evidence based prevention and McGill S, Juker D, Kropf P 1996 Quantitative intramuscular myoelectric Clinical Biomechanics (Bristol, Avon) 1 1 : 1 70-172 activity of quadratus lumborum during a wide variety of tasks. rehabilitation. Human Kinetics Publishers, Champaign, illinois

Indahl A, Kaigle A, Reikeras 0, Holm S 1999 Sacroiliac joint musculature. Journal of Spinal Disorders 12: 325-330

involvement in activation of the porcine spinal and gluteal

Macintosh J E, Bogduk N 1986 The detailed biomechanics of the Magnusson M, Aleksiev A, Wilder D, Pope M, Spratt K, Lee S 1996 back pain. European Spine Journal S: 23-35 lumbar multifidus. Clinical Biomechanics 1: 205-231

Janda V 1978 Muscles, central nervous motor regulation and back problems. In: Korr 1 M (ed) The neurobiologic mechanisms in manipulative therapy. Plenum Press, New York, pp 27-41

Unexpected load and asymmetric posture as etiologic factors in low

Jenison R 1997 On acoustic information for motion. Ecological Johansson R S, Westling G 1988 Programmed and triggered actions to Research 71: 72-86 rapid load changes during precision grip. Experimental Brain Psychology 9: 131-151

Marsden C D, Merton P A, Morton H B 1977 AntiCipatory postural 275: 47P-48P

responses in the human subject. Journal of Physiology (London)

Massion J 1992 Movement, posture and equilibrium: interaction and Matre D A, Sinkjaer T, Svensson P, Arendt-Nielsen L 1998 Experimental Mittelstaedt H 1996 Somatic graviception. Biological Psychology 42: 53-74 Mori S, Iwakiri H, Homma Y, Yokoyama T, Matsuyama K 1995 muscle pain increases the human stretch reflex. Pain 75: 331-339 coordination. Progress in Neurobiology 38: 35-56

Johansson H, Sjolander P, Sojka P 1991 A sensory role for the cruciate

Kaigle A M, Holm S H, Hansson T H 1995 Experimental instability in Kaigle A M, Wessberg P, Hansson T H 1998 Muscular and kinematic Spinal Disorder 1 1 : 163-174 the lumbar spine. Spine 20: 421-430

ligaments. Clinical Orthopaedic and Related Research 268: 161-178

behavior of the lumbar spine during flexion-extension. Journal of

Neuroanatomical and neurophysiological bases of postural control.

In: Fahn S, Hallet M, Leders H, Marsden C D (eds) Negative motoro Moseley G L, Hodges P W, Gandevia S C 2002 Deep and superficial fibres of multifidus are differentially active during voluntary arm Moseley G L, Hodges P W, Gandevia S C 2001a Attention demand, anxiety and acute pain cause differential effects on postural movements. Spine 27: E29-36 phenomena. Lippincott-Raven, Philadelphia

Kelso J A S 1984 Phase transitions and critical behaviour in human Integrative, and Comparative Physiology 15: RI000-I004

bimanual coordination. American Journal of Physiology: Regulatory,

Keshner E A, Allum J H J 1990 Muscle activation patterns coordinating

postural stability from head to foot. In: Winters J M, Woo S L-Y (eds)



activation of the abdominal muscles in humans. Abstracts - Society for Neuroscience 304-31 2 Moseley G L , Hodges P W, Gandevia S C 2001b External perturbation to the trunk is associated with differential activity of the deep and superficial fibres of lumbar multifidus. In: Mooney A V, Gracovetsky V et al (eds) Fourth Interdisciplinary World Congress on Low Back and Pelvic Pain, Montreal, Canada, pp 241-242 Moseley G L, Hodges P W, Gandevia S C 2002 Deep and superficial fibers of lumbar multifidus are differentially active during voluntary arm movements. Spine 27: E29-36 Moseley G L, Hodges P W, Gandevia S C 2003 External perturbation of the trunk in standing humans results in differential activity of components of the medial back muscles. Journal of Physiology 547: 581-587 Mouchnino L, Aurenty R, Massion J, Pedotti A 1990 Coordinated control of posture and equilibrium during leg movement. In: Brandt T, Paulus W, Bles W, Dieterich M, Krafczyk S, Straube A (eds) Disorders of posture and gait. Georg Thieme, Stuttgart, pp 68-71 Mouchnino L, Aurenty R, Massion J, Pedotti A 1993 Is the trunk a reference frame for calculating leg position? Neuroreport 4: 1 25-127 Myriknas S E, Beith I 0, Harrison P J 2000 Stretch reflexes in the rectus abdominis muscle in man. Experimental Physiology 85: 445-450 Nade S, Newbold P J, Straface S F 1987 The effects of direction and acceleration of movement of the knee joint of the dog on medial articular nerve discharge. Journal of Physiology 388: 505-519 Nitz A J, Peck 0 1986a Comparison of muscle spindle concentrations in large and small human epaxial muscles acting in parallel combinations. American Surgeon 52: 273-277 Nitz A J, Peck 0 1986b Comparison of muscle spindle concentrations in large and small human epaxial muscles acting in parallel combinations. American Surgeon 52: 273-277 Oddsson L I, Persson T, Cresswell A G, Thorstensson A 1999 Interaction between voluntary and postural motor commands during perturbed lifting. Spine 24: 545-552 Panjabi M M 1992a The stabilising system of the spine. II: Neutral zone and instability hypothesis. Journal of Spinal Disorders 5: 390-397 Panjabi M M 1992b The stabilizing system of the spine. I: Function, dysfunction, adaptation, and enhancement. Journal of Spinal Disorders 5: 383-389 Panjabi M M, Abumi K, Duranceau J, Oxland T 1989 Spinal stability and intersegmental muscle forces: a biomechanical model. Spine 14: 194-200 Pedersen J, Sjolander P, Wenngren B I, Johansson H 1997 Increased intramuscular concentration of bradykinin increases the static fusimotor drive to muscle spindles in neck muscles of the cat. Pain 70: 83-91 Perry J 1992 Gait analysis: normal and pathological function. SLACK Incorporated, Thorofare, New Jersey Plassman B L, Gandevia S C 1989 Comparison of human motor cortical projections to abdominal muscles and intrinsic muscles of the hand. Experimental Brain Research 78: 301-308 Radebold A, Cholewicki J, Panjabi M M, Patel T C 2000 Muscle response pattern to sudden trunk loading in healthy individuals and in patients with chronic low back pain. Spine 25: 947-954 Radebold A, Cholewicki J, Polzhofer G K, Greene H S 2001 Impaired postural control of the lumbar spine is associated with delayed muscle response times in patients with chronic idiopathic low back Rantanen 1, Hurme M, Falck B et al 1993 The lumbar multifidus muscle five years after surgery for a lumbar intervertebral disc herniation. Spine 18: 568-574 Richardson C A, Jull G A, Hodges P W, Hides J A 1999 Therapeutic exercise for spinal segmental stabilisation in low back pain: scientific basis and clinical approach. Churchill Livingstone, Edinburgh pain. Spine 26: 724-730

Richardson C A, Snijders C J, Hides J A, Damen L, Pas M S, Storm J 2002 The relation between the transversus abdominis muscles, sacroiliac joint mechanics, and low back pain. Spine 27: 399-405 Roberts S, Eisenstein S M, Menage J, Evans E H, Ashton I K 1995 Mechanoreceptors in intervertebral discs: morphology, distribution, and neuropeptides. Spine 20: 2645-2651 Roy S H, DeLuca C J, Casavant 0 A 1989 Lumbar muscle fatigue and chronic low back pain. Spine 14: 992-1001 Sahrman S 2002 Diagnosis and treatment of movement impairment syndromes. Mosby, St Louis Sapsford R R, Hodges P W 2001 Contraction of the pelvic floor muscles during abdominal maneuvers. Archives of Physical Medicine and Rehabilitation 82: 1081-1088 Sapsford R R, Hodges P W, Richardson C A, Cooper 0 H, Markwell S J, JuLl G A 2001 Co-activation of the abdominal and pelvic floor muscles during voluntary exercises. Neurourology and Urodynarnics 20: 31-42 Saunders S, Rath 0, Hodges P W 2002 Respiratory and postural activation of the trunk muscles changes with mode and speed of locomotion. Gait and Posture, in press 2004 Schmidt R A, Lee T 0 1999 Motor control and learning: a behavioural emphaSiS. Human Kinetics, Champaign, Illinois Shirado 0, Ito T, Kaneda K, Strax T E 1995 Flexion-relaxation phenomenon in the back muscles: a comparative study between healthy subjects and patients with chronic low back pain. American Journal of PhYSical Medicine Rehabilitation 74: 139-144 Siddall P J, Cousins M J 1995 Pain mechanisms and management: an update. Clinical and Experimental Pharmacology and Physiology 22: 679-688 Sihvonen T, Lindgren K A, Airaksinen 0, Manninen H 1997 Movement disturbances of the lumbar spine and abnormal back muscle electromyographic findings in recurrent low back pain. Spine 22: 289-295 Sni ders C J, Vleeming A, Stoeckart R 1993 Transfer of lumbosacral load j to iliac bones and legs. 1: Biomechanics of self bracing of the sacroiliac joints and its significance for treatment and exercise. Clinical Biomechanics 8: 285-294 Snijders C J, Vleeming A, Stoeckart R, Mens J M A, Kleinrensink G J 1995 Biomechanical modelling of sacroiliac joint stability in different postures. Spine: State of the Art Reviews 9: 419-432 Solomonow M, Zhou B H, Harris M, Lu Y, Baratta R V 1998 The ligamento-muscular stabilizing system of the spine. Spine 23: 2552-2562 Solomonow M, Zhou B H, Baratta R V, Lu Y, Harris M 1999 Biomechanics of increased exposure to lumbar injury caused by cyclic loading: 1 : Loss o f reflexive muscular stabilization. Spine 24: 2426-2434 Spencer J 0, Hayes K C, Alexander I J 1984 Knee joint effusion and quadriceps reflex inhibition in man. Archives of Physical Medicine and Rehabilitation 65: 171-177 Stokes M, Young A 1984 The contribution of reflex inhibition to arthrogenous muscle weakness. Clinical Science 67: 7-14 Suzuki N, Ohe K, Inoue H 1977 The strength of abdominal and back muscles in patients with low back pain. Central Japanese Journal of Orthopaedics and Traumatology 20: 332-334 Svensson P, Arendt-Nielsen L, Houe L 1995 Sensory-motor interactions of human experimental unilateral jaw muscle pain: a quantitative analysis. Pain 64: 241-249 Taimela S, Kujala U M 1992 Reaction times with reference to musculoskeletal complaints in adolescence. Perceptual and Motor Skills 75: 1075-1082 Tairnela S, Kankaanpaa M, Luoto S 1999 The effect of lumbar fatigue on the ability to sense a change in lumbar position: a controlled study. Spine 24: 1322-1327 Takala E, Viikari-Juntura E 2000 Do functional tests predict low back pain? Spine 25: 2126-2132

Motor control of t h e t r u n k


Taub E, Berman A J 1968 Movement and learning in the absence of sensory feedback. In: Freedman S J (ed) The neurophysiology of spatially oriented behaviour. Dorsey Press, Homewood, illinois Thorstensson A, Arvidson A 1982 Trunk muscle strength and low 69-75 Urquhart 0 M, Hodges P W, Story I, Barker P J, Briggs C A 2001 Regional morphology of transversus abdominis and obliquus internus abdominis. Proceedings of the Biennial Congress of Musculoskeletal Physiotherapists of Australia, Adelaide, Australia Urquhart 0, Story I, Hodges P 2002 Transversus abdominis recruitment in trunk rotation. International Physiotherapy Conference, Sydney, Australia Valeriani M, Restuccia 0, Oi Lazzaro V et al 1999 Inhibition of the human primary motor area by painful heat stimulation of the skin. Clinical Neurophysiology 110: 1475-1480 Vleeming A, Pool-Goudzwaard A L, Stoeckart R, vanWingerden J-p, Snijders C J 1995 The posterior layer of the thoracolumbar fascia: its function in load transfer from spine to legs. Spine 20: 753-758 Wilder 0 G, Aleksiev A R, Magnusson M L, Pope M H, Spratt K F, Goel V K 1996 Muscular response to sudden load: a tool to evaluate fatigue and rehabilitation. Spine 21: 2621>-2639

Wilke H J, Wolf S, Claes L E, Arand M, Wiesend A 1995 Stability increase of the lumbar spine with different muscle groups: a biomechanical in vitro study. Spine 20: 192-198 Williams P L, Warwick R, Dyson M, Bannister L H, (eds) 1989 Gray's Anatomy. Churchill Livingstone, London Winter 0 A, Patla A E, Prince F, Ishac M, Gielo-Perczak K 1998 Stiffness control of balance in quiet standing. Journal of Neurophysiology 80: 1211-1221 Wolf S L, Basmajian J V 1977 Assessment of paraspinal electromyographic activity in normal subjects and chronic low back pain patients using a muscle biofeedback device. In: Asmussen E, Jorgensen K (eds) Biomechanics IV B. University Park Press, Baltimore, pp 319-324 Zattara M, Bouisset S 1986 Chronometric analysis of the posturo-kinetic programming of voluntary movement. Journal of Motor Behaviour 18: 215-223 Zedka M, Prochazka A 1997 Phasic activity in the human erector spinae during repetitive hand movements. Journal of Physiology 504: 727-734 Zedka M, Prochazka A, Knight B, Gillard 0, Gauthier M 1999 Voluntary and reflex control of human back muscles during induced pain. Journal of Physiology 520: 591--604

back pain. Scandinavian Journal of Rehabilitation Medicine 14:


Chapter 11

The lumbar fasciae and segmental control
P. J. Barker, C. A. Briggs

141 141 141 142 142

Anatomy and biomechanics

Anterior layer of lumbar fascia Middle layer of lumbar fascia Fibre orientation
142 142 143

Bony and ligamentous attachments Features and stiffness Muscle attachments

Tensile effects of muscle attachments Posterior layer of lumbar fascia Fibre orientation
144 144 145 144


Bony and ligamentous attachments Features and stiffness Muscle attachments


T he middle and posterior layers of lumbar fasciae encapsu­ late the paraspinal muscles and provide attachment for muscles converging from the back, limbs and abdominal wall. It has been proposed that these fasciae support the lumbar spine and sacroiliac joint via several mechanisms. T his chapter presents current evidence from anatomical, biomechanical, electromyographic (EMG) intra-abdominal (lAP) and intramuscular pressure studies. It incorporates these with proposed functions of fasciae and in particular with models of segmental control. T he magnitude of forces involved and roles in different planes are discussed, with reference to directions for future research and low back pain management.

Features of attached muscle regions Tensile effects of muscle attachments Segmental control
146 146 147

145 145

Comparative features of the middle and posterior layers of the lumbar fasciae Related muscles
147 147 147 148 148 148

Attachments and classification General EMG activity Local regional EMG activity Global regional EMG activity

Biomechanical roles of the lumbar fasciae Longitudinal tension generation Hydraulic amplifier effect Lumbar segmental control Sacroiliac stability Proprioception Planar stability
149 149 150 150 150 150 151 149 148 148 148

Load transfer across the midline

Magnitude of segmental forces Coronal stability Sagittal stability Fascial disruption Conclusion

T he lumbar fasciae are arranged in three layers. The ante­ rior layer (ALF) is thin and membranous while the middle and posterior layers (MLF, PLF) are more fibrous. The latter two attach to lumbar transverse and spinous processes (respectively), collectively enclosing the paraspinal mus­ cles. All three layers meet and fuse at the lateral raphe, between the twelfth rib and iliac crest (Farfan 1995). Attachments at this raphe include fascicles from transver­ sus abdominis (TrA), internal oblique (10) and external oblique (EO) as well as latissimus dorsi (LD) (Barker et al 2004, Bogduk & Macintosh 1984, Bogduk et al 1998, Tesh 1986, Vleeming et a11995) (Fig. 11.1). Lumbar fasciae are also termed 'thoracolumbar ' fasciae, although only the posterior layer extends above the level of the twelfth rib and correctly deserves this name. Even 'fas­ cia' may be an inappropriate classification for these tissues (Bogduk 1997, Gallaudet 1931), since the MLF and PLF blend medially with vertebral ligaments and form aponeu­ rotic attachments for TrA and LD, so might also be consid­ ered ligamentous or tendinous (Bogduk 1997).
Anterior layer of lumbar fascia

Transverse stability

The anterior layer of lumbar fascia (ALF) covers quadratus lumborum (QL), joins the MLF laterally at the lateral raphe





The lumbar fasciae in cross-section at L4 and L2. Note 10's attachment to the lateral raphe below L3 and EO's attach ment to

it above L3. Reprod uced from Barker and Briggs Key: EO Mf


Spine 24 (17) : 1757-1764 with permission from Lippin cott, Williams Et Wilkins.

external oblique; 10


internal oblique; TrA


transversus abdominis; LD


l atissimus dorsi; QL


quad ratus l u mborum; Ps



mul tifidus; ALF

anterior lumbar fascia; MLF

middle l u m ba r fascia; PLF

posterior l u mbar fascia.

and inserts medially on the anterior surface of each lumbar transverse process. It is thin (0.1 mm), membranous (Barker et al 2004b) and may blend with the fascia over psoas laterally. The ALF displays thickenings superiorly and laterally. The lateral arcuate ligament is the superior thickening, pro­ viding attachment for the diaphragm and covering the upper part of QL. A second thickening passes vertically between the tip of the twelfth rib and the iliac crest. The remainder of the ALF lacks fibres and its capacity for tensile transmission appears to be minimal.
Middle layer of lumbar fascia
Bony and ligamentous attachments

with fascicles from the mid-region of TrA (Barker et al 2004b, Urquhart et a12004). At the lateral raphe, a few fibres of the MLF may be reflected posteriorly to join the deep lamina of the PLF, encircling the lateral border of erector spinae (Tesh et al 1987). Since fibre orientation indicates the directional stiffness of a tissue (Hukins 1984, 1985; Minns et al 1973), the MLF is likely to be stiffer transversely.
Features and stiffness

The width of the MLF, from transverse processes to lateral raphe, is only 2-3 cm, the aponeurosis of TrA extending

The middle layer of lumbar fascia (MLF) arises from the iliac crest and posterior iliolumbar ligament, attaching superiorly to the medial part of the twelfth rib and lumbo­ costal ligament (Bogduk & Macintosh 1984, Williams et al 1995). Here, QL is tightly enclosed between the lumbocostal ligament and lateral arcuate ligament (Poirier 1901). Medially, the MLF attaches to the outer edge of each lum­ bar transverse process (Barker et a12004b, Breathnach 1965, Sharpey et al 1867, Tesh et al 1987) and the intertransverse ligaments. Laterally, the MLF has only muscular attach­ ments, of which the most extensive is to TrA (Fig. 11.2).
Fibre orientation

Fibres of the MLF radiate laterally from the tips of lumbar transverse processes. Superolateral fibres are short (-2 cm), angled up to 30 degrees above the horizontal before joining inferolateral fibres from the transverse process above, to form fibrous 'arches' between the processes (Barker et al 2004b, Tesh et al 1987, Testut & Latarjet 1948) (see Fig. 11.2). The majority of fibres are directed inferolaterally (approxi­ mately 10-25 degrees below horizontal) and are continuous



The middle layer of lumbar fascia. Note the thi.<;k

attachments of the MLF to tran sverse p rocesses, the extensive TrA aponeu rosis a n d that MLF fibres are angled m o re towards horizon­ tal inferiorly. Reproduced with permission from Barker et al 2004b. Key: TrA

transversus abdominis; MLF


middle l u mbar fascia.

The lumbar fasciae and segmental control


beyond this for 5-6 cm (Barker et al 2004a, 2004b) (see Fig. 11.2). The average thickness of the MLF varies greatly, being more than 50% thicker at its attachments to the transverse processes than between them (0.62:0.40 mm; Barker et al 2004b). These attachments are aligned primarily horizon­ tally and may avulse the transverse process tips during contraction of TrA (Marshall 2001). Avulsion of the processes has been observed in dissections by traction on the MLF, confirming its capacity to transmit loads to the vertebral column (Barker et al 2004b). In unembalmed cadavers, .transverse tensile loads of up to 50 N have been applied to TrA without the muscle or MLF tearing; how­ ever, their maximum capacity is likely to be greater than 50 N (Barker et a12004a). Although mechanical testing of the MLF on isolated segments is difficult (Tesh 1986), early anatomical texts referred to it as the strongest layer of lumbar fascia (Davies-Colley 1894, Sharpey et al 1867) and limited in vitro tests support this. In an unembalmed cadaver, Tesh (1986, 1987) inserted a balloon confined (by rigid horizon­ tal dividers) to the lumbar region of the abdomen, then applied lateral flexion torques while varying the pressure in the balloon. Increasing the pressure (from 60 to 120 mmHg) was noted to right the trunk against applied lat­ eral flexion force, requiring a proportional increase in restraining force necessary to retain lateral flexion. T his force was up to 14.5 Nm or equivalent to 40% of the moment produced by body weight in extreme lateral flex­ ion. If the position was sustained and both pressure and lateral flexion torque removed, then the pressure re­ raised, the cadaver moved back to the neutral position. The force was attributed to tension generation in the MLF (Tesh 1986, Tesh et al 1987). The MLF possesses thickenings via which it may trans­ mit loads to the vertebral column. Although its stiffness has not been quantified, its tensile capacity is likely to be greater horizontally. The capacity of the MLF to resist loads in the coronal plane, at end of range lateral flexion, may be up to 14.5 Nm.

Tensile effects of muscle attachments

Raster photography, a technique to indicate tensile trans­ mission between muscles and fascia by comparing move­ ment of fascial markers against a reference grid, was described by Vleeming et al (1995). Tension was applied to fascial attachments to simulate the effect of muscle contraction. A similar technique was employed to determine the area of fascial movement following tensile loading of the lateral abdominal muscles and MLF in unembalmed cadavers (Barker et al 2004a). Tension on the attachment of TrA dis­ placed more than double the area of fasciae than tension on 10 or EO (Fig. 11.3) and resisted higher applied tension before failure. A strain gauge inserted into the MLF and PLF at their vertebral (13) attachments indicated tension applied to TrA was transmitted more effectively to the MLF, its ten­ sile force being almost twice that of the PLF and transmitted direct to vertebrae rather than across the midline (Barker et a12004a) (see Fig. 11.3). Tesh's work (Tesh 1986) indicates the MLF is also most likely to transmit tensile forces generated from within TrA, since the intra-abdominal balloon pressure was sustained even when the PLF was incised. Mathemati­ cal calculations (Barker et al 2004a) indicate that during maximum contraction, the MLF may withstand a transverse hoop tension of 48 N per segment. TrA therefore appears to be the only muscle that can transmit tension via the MLF to all lumbar vertebrae, its fascicles being attached and well aligned with fibres of the MLF throughout this region. TrA is the primary muscle attachment of the MLF, which in turn appears well struc­ tured to operate as this muscle's principal aponeurosis of origin.

70 60 50

Muscle attachments

40 30 20 10 0 LD TrA GM 10 EO TrA2 102 E02 Posterior layer Middle layer


The MLF is attached to fascicles of TrA, LD (Bogduk & Macintosh 1984), EO and 10 (Barker et a12004a, Bogduk & Macintosh 1984, Vleeming et al 1995). EO attaches to it above the level of the transverse process of L3, 10 below this level and TrA to the full length of the lateral raphe (Barker et a12004a). The attachments of LD, 10 and EO are relatively small and muscular (Gallaudet 1931) and fasci­ cles of 10 and LD oriented almost perpendicular to fibres of the MLF. Other muscles including QL, iliocostalis lum­ borum and the diaphragm have small attachments to the MLF. By contrast, the attachment of TrA is extensive and aponeurotic laterally (Testut & Latarjet 1948), with fasci­ cles being directly continuous with fibres of the MLF (Barker et al 2004b).




Tensile force


Fascial area and tensile forces. The areas of PLF and MLF

displaced and tensile forces developed in fasciae with


N applied

tension to attached muscles. 'Tensile force' indicates the transverse component of tensile force at L3. Reproduced from Barker and Briggs



24 (17): 1757-1764 with
= =

permission from Lippincott,

Williams 8: Wilkins. Key: PLF lum bar fascia; LD

posterior lumbar fascia; MLF


latissimus dorsi; TrA

transversus abdominis; GM

gluteus maximus; 10

internal oblique; EO

external oblique.



Posterior layer of lumbar fascia

The posterior layer of lumbar fascia (PLF) surrounds the paraspinal muscles and consists of two laminae, which are progressively fused below T12 (Barker & Briggs 1999, Bogduk & Macintosh 1984). Figure 11.4 illustrates these two laminae with their fibre directions and attachments.
Bony and ligamentous attachments

The PLF attaches to lumbar and thoracic spinous processes, supraspinous and interspinous ligaments, the posterior superior iliac spine and posterior part of iliac crest as well as the ilium on the opposite side (Bogduk & Macintosh 1984, Tesh et al 1987). It is also reported to be continuous inferiorly with the long dorsal sacroiliac and sacrotuberous ligaments (Vleeming et al 1996, 1995), while its deep lamina attaches superolaterally at each rib angle (Barker & Briggs 1999). Fibres from the superficial lamina of the PLF cross the midline at all lumbar levels (Tesh et al 1987) although midline attachments become less evident below L3 (Bogduk & Macintosh 1984, Vleeming et aI1995). Each interspinous ligament, via its sagittally oriented fibres, attaches the deep lamina of the PLF to the anterior upper border of the spinous process below. Its primary function has been proposed to anchor the PLF to the spine (Aspden et al 1987, Hukins et a11990b) or to limit anterior shear forces on vertebrae (Farfan 1995, Tesh et al 1987). Lamellae within the spinous processes are also aligned in the sagittal plane (Gallios & Japiot 1925). However, the PLF itself approaches the spinous processes from a posterolat­ eral angle, which varies with contraction of the paraspinal muscles (Tesh 1986). The midline lumbar attachments of the PLF are variable and may require paraspinal muscle contraction for effective tensile transmission to the interspinous ligaments and spin­ ous processes.
Fibre orientation



The posterior layer of lumbar fascia. The PLF: muscu­

lar attach ments of deep (left) and superficial (right) laminae. Reproduced from Barker and Briggs Key: See Bi



24 (17): 1757-1764

with permission from Lippincott, Wil l iams Et Wilkins.

splenius cervicis; SPI


serratus posterior inferior;

biceps fem oris; Tz

trapezius; Rh

rhom boids; GM




The superficial lamina of the PLF appears cross-hatched below T 12. This has been attributed to fibres aligning superolaterally with LD and inferolaterally with the con­ tralateral gluteus maximus (GM), (Vleeming et al 1995), although since the laminae are fused, exact origins are dif­ ficult to trace (Bogduk & Macintosh 1984). Serratus poste­ rior inferior and TrA also have fascicles closely oriented with fibres of the PLF (see Fig. 11.4) so its fibre directions may be partly attributed to these muscles (Barker et al 2004a). Fibres in the deep lamina are predominantly super­ olateral (Bogduk & Macintosh 1984). Fibre angles of the PLF are generally measured with the spine in extension and are cited here with reference to a horizontal axis. Barker & Briggs (1999) reported angles and/ or presence of fascial fibres at alternate vertebral lev­ els throughout the PLF in 20 cadavers. Superolateral fibres in the superficial lamina become more transverse in the upper lumbar spine, oriented almost 40 degrees above the

horizontal at L5 and 30 degrees at Ll. At T9 they lie trans­ versely, then are oriented increasingly inferolaterally to become continuous with the rhomboid muscles (see Fig. 11.4). In the lumbar region, inferolateral fibres (20 degrees below horizontal) cross the above fibres to create a hatched appearance (Barker & Briggs 1999). The lattice formed by PLF fibres is thus more Closely ori­ ented towards a horizontal axis. Fibres in the deep lamina above T12 are consistently directed superolaterally at 20 degrees above the horizontal but decrease at T4, above which this lamina becomes more membranous (Barker & Briggs 1999).
Features and stiffness

The average thickness of the PLF (0.52 mm) is comparable with that of the middle layer (0.55 mm; Barker et aI2004b). Thickened fascial bands have been described in the PLF (Bogduk & Macintosh 1984) but regional differences are

The lumbar fasciae and segmental control


only minor (0.56 mm) at the spinous processes compared with between them (0.53 mm; Barker & Briggs 1999). The increase in thickness at vertebral attachments (6%) is slight compared with that observed in the middle layer (55%; Barker et a12004b). Both laminae of the PLF diminish in thickness and fibrosity in the thoracic region, yet the region of the super­ ficial lamina between LO and rhomboid attachments can vary considerably (Barker & Briggs 1999). This lamina ends freely at the upper border of rhomboid minor while the deep lamina remains thin throughout the thoracic region, attaching superiorly to splenius cervicis (Barker & Briggs 1999) and fascia overlying the splenii (Bogduk & Macintosh 1984). The deep lamina thus forms an enclosed compart­ ment that runs the length of the spine, surrounding the paraspinal and splenius muscles. Dye injection studies sup­ port this (Peck et al1986). The width of the PLF, from spinous processes to lateral raphe, is approximately 7 cm (Barker et al 2004a) per side. It lies approximately 7 cm behind the instantaneous axis of rotation for flexion, with almost double the moment arm of most posterior ligaments (Tesh 1986, Tesh et al 1987). Mechanical testing using samples of PLF indicates it may be up to four times stiffer transversely than longitudinally, and three times stiffer transversely than in an oblique (inferolateral) direction (Tesh 1986). Consistent with other connective tissues, stiffness of the PLF increases with defor­ mation (Tesh 1986, Yahia et al1993). Transverse stiffness of samples of PLF has been reported at up to 113 MPa, which is relatively unimpressive com­ pared with other fasciae (Tesh 1986). T his, however, may be an underestimate, since isolated samples tend to be weaker than entire intact tissues with in situ attachments (Adams 1995, Adams & Green 1993). During maximal IAP and with contraction of the paraspinal muscles, the PLF has been estimated mathematically to withstand a maximum trans­ verse ('hoop') tension of 90 N at each segment (Tesh 1986), although subsequent analysis indicates this tension is more correctly attributed to the MLF (Barker et al 2004a). Tests applying sub-failure tension to attached fascicles of TrA indicate the PLF is likely to withstand at least 50 N tension (Barker et a12004a). The numeric stiffness of the PLF is uncertain, but its transverse stiffness is up to four times greater than its lon­ gitudinal stiffness. Small tensile loads applied to muscle attachments are easily transmitted via the lumbar fasciae to vertebrae (Barker et al 2004a) and may play a role in influ­ encing segmental movement around the neutral position.
Muscle attachments

In the thoracic region, the superficial lamina is attached to the rhomboids up to the level of T2 and to the lowest fascicles of trapezius at T11/12. The superficial lamina is also attached to LO, via an oblique aponeurotic junction above the iliac crest, and the contralateral GM, via fibres crossing the midline below L4. Inferiorly it is attached to ipsilateral fascicles of GM between the poste­ rior superior iliac spine and the median sacral crest, at S2-4 levels. T he deep lamina is attached superiorly to the lower border of splenius cervicis. Here, serratus posterior supe­ rior (SPS) overlies the deep lamina without attaching to it and is, in turn, located deep to the rhomboids. SPS thus separates the two laminae of the PLF without attaching to either, while by contrast, at upper lumbar levels, serratus posterior inferior (SPI) is entirely continuous with the deep lamina. T he deep lamina attaches laterally to the mid-region of TrA and to part of EO above L3 and part of 10 below it. This lamina also has variable inferior attach­ ments to the lumbar erector spinae aponeurosis (below L5; Hutchinson & Oall 1994) and the sacrotuberous ligament, which may provide indirect attachment to biceps femoris (Vleeming et al 1989).
Features of attached muscle regions

The PLF receives the same attachments via the lateral raphe as the MLF (see Fig. 11.1A, B) as well as attachments from many extrinsic back muscles. These are illustrated (see Fig. 11.4) from superiorly to inferiorly, using combined findings of dissection and tensile studies (Barker & Briggs 1999, Bogduk & Macintosh 1984, Vleeming et al 1995).

Most attachments to the PLF consist of groups of adjacent fascicles rather than entire muscle bellies, the exceptions being SPI and rhomboid muscles. All attachments have varying fascicle directions, lengths, moment arms and cross-sectional areas and consequently different vectors and capacities for stiffness or torque generation (Bergmark 1989). Attaching fascicles of LO and GM are relatively long, directed superolaterally towards the axilla and inferolater­ ally towards the iliotibial tract, respectively. SPI has shorter fascicles, passing superolaterally toward the lower four ribs. Attaching fascicles of EO and TrA are relatively long and pass anteroinferiorly; those of EO almost vertically to the anterior iliac crest and those of TrA to the rectus sheath. 10's attached fascicles are comparatively short, passing superolaterally to the lower three to four ribs (Williams et al 1995) (see Fig. 11.4). Cross-sectional areas of the muscle attachments vary considerably. LO, TrA, SPI, EO and 10 are generally thin, broad muscles, with fascial attachments of LO and TrA being the most extensive. Bogduk et al (1998) reported fibres of LO had insufficient cross-sectional area to pro­ duce torque via the PLF at the sacroiliac joint (Bogduk et al 1998), but the attachment of GM appears to have a more substantial cross-sectional area (P. J. Barker, unpublished work, 2003).
Tensile effects of muscle attachments

Vleeming et al (1995) applied 50 N tension to attachments of the PLF in embalmed cadavers, reporting that both LO and GM transmitted tension to the posterior layer of




lumbar fascia contralaterally, LD up to 2 cm and GM up to 4 cm past the midline. Traction to 5PI and EO, biceps femoris and trapezius and gluteus medius produced vari­ able, limited or no fascial displacement, respectively. Barker et al (2004a) performed a similar study in unembalmed cadavers using a lower (10 N) applied tension and found greatly increased tensile transmission, with every muscle tested (LD, GM, TrA, 10, EO) producing fascial movement and tensile force (see Fig. 11.3). Most extensive PLF displacement resulted from tension on LD and TrA, bilaterally between T12 and 51, while ten­ sion on GM and 10 caused fascial displacement below L3 and tension on EO above it. Tension on both obliques often produced only unilateral displacement (Barker et al 2004a). Transverse tensile force in the PLF was also found to be greatest when tension was applied to LD and TrA, with up to 50% of applied tensile force being transmitted to the fas­ cia adjacent to the L3 spinous process (Barker et al 2004a) (see Fig. 11.3). Two studies by Tesh (1986) concur that TrA transmits tension to the PLF. During inflation of a lumbar intra­ abdominal balloon, markers on the PLF were noted to move laterally and if incised longitudinally, the edges of the incision sprang apart, indicating the PLF was transmitting tension. Tesh's concurrent CT study indicated that a Valsalva manoeuvre (performed in supine) drew the PLF anteriorly. Contraction of the paraspinal muscles has also been proposed to generate transverse tension in the PLF (Carr et al 1985, Farfan 1973, Gracovetsky et al 1977, Tesh et al 1987), with staining techniques indicating this mecha­ nism may transmit tension to both supraspinous and inter­ spinous ligaments (Tesh 1986). A study of the PLF which perhaps has the greatest implications for segmental stability was also performed by Tesh (1986). T he effects of applying lateral fascial ten­ sion on segmental sagittal rotation were investigated. Lateral tension of 20 N was applied direct to the PLF, via grips to simulate contraction of the lateral abdominal muscles, then segments were loaded in sagittal rotation via distracting their spinous processes. When lateral ten­ sion was applied, resistance to sagittal rotation increased from the onset of loading and throughout early range at all segments (Tesh 1986), so that additional sagittal force (e.g. 20 N) was required to achieve the same spinous process distraction (e.g. 1 mm; Fig. 11.5). Transverse load­ ing of the PLF effectively stiffened the motion segment in the initial stage of flexion, without altering the final stiff­ ness (Tesh 1986). The PLF has extensive attachments to LD and TrA. Tensile and staining tests indicate that every muscle attach­ ment is capable of generating some tension in the PLF and so via it may contribute to spinal stability (Bogduk & Macintosh 1984). Transverse tension in the PLF can increase segmental stiffness to sagittal plane rotation in early range (Tesh 1986).

PLF tension and L3I4 movement (sagitlal)

20N lateral tension on


� .E OJ E
"0 Q)

200 150 100 50 0 0 2

Zero tension on




.S! Ci.
a. «

Zones of






Spinous process distraction (mm)



The effect of PLF tension on sagittal segmental

movement. Typical load deformation cu rve taken from the L3-4 segment d u ri n g sagittal rotatio n . Note that with

20 N

ap plied lat­

eral tension to the PLF, the cu rve moves to the left, its gradient in dicating an increased stiffness in early sagittal displacemen t but n o change at the end of the range. Adapted from Tesh Key: PLF


posterio r l umbar fascia.

Comparative features of the middle and posterior layers of the lumbar fasciae

Both MLF and PLF are capable of transmitting tension from TrA to all lumbar vertebrae, with fibres of the middle layer being directly continuous with TrA fascicles laterally and thickened at their attachments to the transverse processes, medially. From the lateral raphe, the MLF provides a rela­ tively short, direct route to vertebrae that generates almost double the transverse tension of the PLF when tension is applied to TrA. The MLF also appears to be the primary fas­ cial restraint for tension generated within TrA (via an intra­ abdominal balloon) and may resist substantial lateral flexion torques. It is, therefore, well structured to transmit a wide range of tensile loads from TrA, from one or both sides. By contrast, the PLF's multiple fibre orientations indicate its suitability for tensile transmission from several attached muscles. Fibres in the PLF are not consistently oriented with fascicles of TrA and it forms a less direct route from TrA, which has been shown to transmit less tensile force (Barker et a12004a) to vertebrae. Although the PLF is stiffest in the transverse direction, its vertebral attachments are rel­ atively inconsistent, unthickened and mobile in the absence of paraspinal contraction. The MLF may consequently pro­ vide a more efficient route for TrA to influence segmental movement than the PLF. Despite this, application of transverse tension to the PLF has been shown to increase resistance to inner range seg­ mental movement in the sagittal plane (Tesh 1986).

The l umbar fasciae and segmental control



Whether a similar stabilizing effect may be produced via the MLF remains to be demonstrated.
Related muscles
Attachments and classification

T he MLF and PLF thus provide mechanisms via which (primarily) TrA may influence segmental control. T heir effi­ ciency depends on the effectiveness of contraction of TrA, which requires optimum motor control (Hodges & Richardson 1996, McGill & Norman 1993).
Local regional EMG activity

Functionally, spinal muscles may be classified as either 'global' or 'local' (Bergmark 1989). Global muscles attaching to the spine, including GM or LD, are generally superficial and responsible for generating large torques, whereas local muscles such as multifidus are deeper, with segmental attachments that generate small torques but are more important in influencing intersegmental movement. Although not originally classified (Bergmark 1989), TrA's anatomy (Barker et al 2004b, Urquhart et a12004), and EMG activity (Hodges & Richardson 1997b) are consistent with a local function (Richardson et aI1999). T he lumbar fasciae are related to muscles from both groups as well as some that are classified as part local and part global. T hese include QL, 10 and certain regions of the erector spinae (Bergmark 1989). 10 was considered to be a partly local muscle due to its attachment to the lumbar fasciae (Bergmark 1989) and so EO might similarly be allocated to both categories. Both obliques are, however, primarily attached to the ribcage and pelvis, so better suited to act as global trunk rotators. Local and global muscles may respond differently to low back injury, with wasting and/or dysfunction documented in the former and excessive levels of co-contraction pro­ posed to develop in the latter (Richardson et aI1999).
General EMG activity

In healthy subjects, collective EMG findings reveal that local muscles such as TrA and the deep fascicles of multi­ fidus are active prior to the prime movers for limb move­ ments, in all directions and at various speeds (Hodges & Richardson 1997a, 1997b, 1998, Moseley et al 2002). T heir onset does not vary with the direction of limb movement, in contrast with onsets of adjacent muscles (Hodges & Richardson 1996, Moseley et al 2002) and their subsequent contraction is more tonic (sustained and submaximal) in nature (Hodges et al 1999, Richardson et aI1999). From these findings, pre-contraction of TrA has been hypothesized to limit excess intersegmental movement occurring either via the lumbar fasciae or IAP generation (Hodges & Richardson 1997b). Tension from TrA may be transmitted via the MLF and PLF to the processes of lum­ bar vertebrae, 'anchoring' them during perturbations to the spine to help prevent excessive movement and potential segmental injury (Hodges & Richardson 1997b). Lumbar multifidus is relatively bulky (Macintosh et al 1986), so might also influence intersegmental movement by increas­ ing tension in the overlying PLF. Only some of its fascicles, however, demonstrate an early onset of contraction, so its effect on vertebrae via the PLF is likely to be less efficient at limiting segmental movement than its direct compressive action (Hodges & Richardson 1996, Moseley et aI2002).

Regions of muscles that have different fascial attachments may also have different functions. To clarify the effect of each muscle, EMG activity within the relevant (attached or enclosed) muscle region must be considered. T he middle fascicles of TrA, which originate between the iliac crest and lower border of the costal margin, are those typically recorded from in EMG studies (Cresswell et al 1992, Hodges & Richardson 1997b, 1998, Hodges et al 1999). T he contraction onset of TrA is noted to be delayed relative to the onset of the prime mover in sub­ jects with low back pain (Hodges et al 2001) and the subsequent contraction becomes more phasic in nature (Hodges & Richardson 1996, Hodges et al 2001). TrA's contraction prior to limb movement has invariably been reported unilaterally, although studies performed on dif­ ferent sides and during bilateral limb movement consis­ tently show this feature (Hodges & Richardson 1997b, Hodges et al 1999), indicating the contraction may occur bilaterally. Additional activity in middle fascicles of TrA has been noted during trunk and pelvic rotation, with greater (uni­ lateral) activity in ipsilateral trunk rotation and opposite pelvic rotation (Cresswell et aI1992). TrA's lowest fascicles may behave in a more tonic fashion than its middle fascicles throughout limb movements (Hodges et aI1999). Only the posterior fascicles of 10 and EO attach to ver­ tebrae via the MLF and PLF. T hese fascicles are proposed to have the greatest capacity for stiffness generation (Bergmark 1989, Mirka et al 1997). However, electrode placement in most EMG studies of the obliques allows recording from fascicles either at or anterior to the mid­ axillary line (Carman et al 1972, Davis & Mirka 2000, Hodges & Richardson 1997b, Mirka et a11997), above the region where EO's fascicles attach to the fasciae and ante­ rior to fascicles of 10 with a fascial attachment. T he pro­ posed local function of these fascicles thus remains to be clarified. T he deep fascicles of lumbar multifidus have been shown to demonstrate an early onset of contraction prior to prime movers for limb movements and are thought to influence vertebral movement by their direct (compressive) action (Moseley et al 2002). In addition, multifidus may sometimes be active bilaterally during rotation (Donisch & Basmajian 1972). Via both the MLF and PLF, an isolated, submaximal con­ traction of TrA may subtly increase stiffness of lumbar seg­ ments prior to trunk perturbations. Bilateral contraction of TrA is required for a symmetric fascial influence on segmental movement.



Global regional EMG activity

Anterior fascicles of EO and IO appear to behave predomi­ nantly in a global fashion. Each has an onset of contraction more specific to the direction of limb movement, with sub­ sequent activity being more phasic (Hodges et al 1999, Richardson et al 1999). Subjects suffering low back pain may recruit fascicles of 10 differently during drawing in of the abdominal wall (O'Sullivan et al 1997a). 10 has also been noted to demonstrate an early onset of contraction prior to some limb movements, but not as early or as con­ sistently as TrA (Hodges & Richardson 1997c, Richardson et al 1999). 10 and EO are also recruited unilaterally as ago­ nists for trunk rotation (Carman et al 1972) and bilaterally for flexion or to stabilize leg movements (Floyd & Silver 1950). During gait and trunk rotation, fascicles of LD and GM that attach to the PLF are reported to contract contralater­ ally in an alternating phasic fashion (Mooney et al 2001) and may generate oblique tension across the PLF (Vleeming et al 1995). The paraspinal muscles may also generate ten­ sion in the PLF during contraction, for example during resisted extension or anti-gravity flexion (Donisch & Basmajian 1972, Floyd & Silver 1955). While tension in the MLF is primarily influenced by contraction of TrA, the PLF may be more influenced by con­ traction of global, phasic muscles. This puts the MLF at an advantage for segmental stability.
Biomechanical roles of the lumbar fasciae
Longitudinal tension generation

Gracovetsky et aI1977). Intramuscular pressure recordings (Styf & Lysell 1987) support restriction by fascia, with mathematical analysis (based on behaviour of non-vis­ coelastic materials) indicating it may increase the effective­ ness of the paraspinals by up to 30% (Hukins et aI1990a).
Lumbar segmental control

The PLF has the largest moment arm of all extensor tissues (Tesh 1986) and has been proposed to oppose flexion moments in several ways. Although initially thought to generate passive longitudinal tension (Gracovetsky et al 1981), its fibres are not well oriented for this, possessing a relatively small vertical component (Barker & Briggs 1999, Bogduk & Macintosh 1984, McGill & Norman 1988, Tesh et a11987, Vleeming et aI1995). Similar to a lattice, lateral ten­ sion generated by abdominal wall muscles might increase longitudinal tension in the PLF. T his has been simulated in cadavers and a small amount of lumbar extension noted (Fairbank & O'Brien 1980). Initially the axial 'gain' was pre­ dicted mathematically to be up to 5:1 (Gracovetsky et al 1985), but subsequent studies indicate gains of less than 1:1, even if the effects of lumbar flexion on fibre angles are incorporated (Tesh et aI1987). Fasciae appear to contribute to spinal support more effectively by generating transverse than longitudinal tension.
Hydraulic amplifier effect

From the neutral position of the lumbar spine, each seg­ ment can move within a region of minimal resistance from surrounding tissues, known as the neutral zone (Panjabi 1992b). Control of this zone is thought to provide a better indication of spinal instability than physiologic range of movement (Panjabi 1992b, Panjabi et aI1989). Surrounding tissues can be classified into active, passive and neural con­ trol subsystems, and deficiency in any one has been pro­ posed to permit excess movement, resulting in pain and disability (Panjabi 1992a). The MLF and PLF may limit lumbar segmental move­ ment by transmitting tension from contraction of TrA (Hodges & Richardson 1997b). Tesh's (1986) work (Tesh 1986) supports this theory, since applying lateral tension to the PLF effectively moved the segmental load deformation curve to the left, reducing the neutral zone (see Fig. 11.5). Although rarely performed, lumbar fasciectomy may result in a sensation of instability (S. N. Bell, personal communi­ cation, 2002) that is consistent with this proposal. Lateral tension on the MLF was not tested but might be predicted to have a similar effect on segmental neutral zone movement and perhaps be more effective at transmitting tension from TrA in the transverse plane (Barker et al 2004a). T he MLF and PLF are passive components in this mechanism, reliant on a relatively isolated bilateral, sym­ metrical contraction of TrA (and/or multifidus) to be effec­ tive. EMG findings collectively indicate that, in vivo, such a fascial mechanism is likely to precede trunk perturbations from the neutral position in the sagittal and coronal planes (Cresswell et al 1992, Hodges & Richardson 1997b, 1998, Moseley et aI2002). Changes in onset of TrA observed with low back pain (Hodges & Richardson 1996, 1998, Hodges et a12001) sup­ port this model of segmental stability, as do the results of rehabilitation aimed at retraining its motor control Gull & Richardson 2000). Decreases in pain and functional disabil­ ity have been noted in patients with spondylolysis and spondylolisthesis as well as reduced recurrence of injury in patients with first episode low back pain. Improvements following motor control retraining have been sustained for more than 2 years (Hides et a12001, O'Sullivan et aI1997b). All of this provides considerable support for a role of TrA (and the MLF and/or PLF) in segmental control.
Load transfer across the midline

By enclosing paraspinal muscles and restricting radial expansion, the PLF may increase longitudinal tension in these muscles, enhancing their contraction and the exten­ sion moment generated by them. This is known as the hydraulic amplifier effect (Aspden 1992, Farfan 1973,

noting that tension on LD and GM caused displacement of markers on the PLF bilaterally, Vleeming et al (1995) pro­ posed that this layer may assist load transfer across the midline, particularly during activities involving 'contralat-

The l umbar fasciae and segmental control


eraIlimb extension or trunk rotation, such as swimming or walking. Loose midline attachments of the PLF may enable some forces to be distributed across, rather than entirely to, the lumbar spine and sacroiliac joint. T his proposal is sup­ ported by findings of tension studies (Barker et al 2004a, Vleeming et a11995) and the co-contraction of contralateral GM and LD during gait and trunk rotation (Mooney et al 2001). Relatively low tensile loading has been noted to pro­ duce contralateral tension in the PLF (Barker et al 2004a, Vleeming et al 1995), indicating this layer has a greater capacity for contralateral tensile transmission tension, but potentially a reduced capacity for restraining movement of vertebrae, via the spinous processes.
Sacroiliac stability

fashion to provide additional posterior compression via the PLF.

The PLF lies directly behind the sacroiliac joint (SIJ), so ten­ sion in it can contribute to active force compression at the SIJ (Vleeming et aI1995). A functional relationship between the SIJ and PLF (with its attached muscles; particularly GM) is supported by several studies. BiomeG1.anical analysis of LD indicates that it may pro­ duce a limited effect (via the PLF) at the SIJ (Bogduk et al 1998). Tensile testing studies similarly indicate the effects of LD and TrA may only extend via the PLF to fascial markers at Sl (Barker et al 2004a, Vleeming et aI1995). By contrast, the fascial attachment of GM is thicker, more proximal to the SIJ and oriented perpendicular to its articular compo­ nents (Dijkstra et aI1989), displacing markers as low as S3 (Barker et aI2004a). Surface EMG studies indicate GM displays a greater sig­ nal amplitude than LD during gait and trunk rotation (Mooney et al 2001) and contraction of GM corresponds with a greater increase in SIJ stiffness (Wingerden et al 2001). T he level of attachment of GM and the PLF also cor­ responds with the level at which pelvic belts are placed (Vleeming et aI1992) to brace the SIJ for effective relief of peripartum pain (Mens et aI1996). However, management regimes for sacroiliac dysfunction based on associations between GM and LD have given varied results (Mens et al 2000, Mooney et al 2001), while TrA is emerging to play a more important role in SIJ stiffness. Richardson et al (2002) proposed that compression of the SIJ is produced by TrA's anterior iliac attachments, which are direct rather than via the lumbar fasciae. Pelvic belt placement, previously thought to simulate the effect of GM and the PLF posteriorly, might more correctly simulate the effect of TrA's fascicles anteriorly (Richardson et al 2002). T he deep fascicles of multifidus also appear appro­ priate to contribute to stability of the SIJ via their anatom­ ical features (Macintosh et al 1986) and EMG behaviour (Moseley et aI2002). It is, therefore, increasingly evident that the lumbar fas­ ciae and their attachments may be less important in sus­ taining SIJ stability than the anterior fibres of TrA. During certain activities, GM and LD may be recruited in a phasic

T he PLF and MLF may play a proprioceptive role in lumbar stability (Barker & Briggs 1999). Attached to ligaments and muscles as well as containing mechanoreceptors (Yahia et al 1992), they are closely related with each of Panjabi's pas­ sive, active and neural subsystems (Panjabi 1992a). T he fas­ ciae may form functional interfaces between these structures at a segmental level. Feedback from mechanore­ ceptors on the status of tension in related muscles and liga­ ments may be incorporated by the neural subsystem and tension in muscles modified to help prevent excess seg­ mental movement (Panjabi 1992a). If this mechanism is disrupted, proprioceptive feedback will be altered. A reduction in trunk proprioception and resultant sensation of instability following lumbar (PLF) fasciectomy (S. N. Bell, personal communication, 2002) is consistent with this. Histological studies report innervation of the PLF to be deficient in patients with chronic low back pain (Bednar et aI1995), which might similarly reduce pro­ prioceptive feedback and segmental control. Proprioceptive deficiency is in keeping with reports of patients with low back pain having difficulty achieving preferential contrac­ tion of TrA (Richardson et al 1999) and highlights the importance of persisting with motor control retraining and incorporation into functional tasks (O'Sullivan et aI1997b) to avoid recurrence of injury.
Magnitude of segmental forces

Just two degrees of axial rotation has been reported to pro­ duce microtrauma of the intervertebral disc (Farfan et al 1970) and very small amounts (3%) of muscle contraction have been predicted capable of restoring segmental stabil­ ity (Cholewicki & McGill 1996). Although limitation of neu­ tral zone movement is now recognized as important in injury prevention (Cholewicki & McGill 1996, Panjabi 1992b), the stabilizing effects of TrA and/ or lumbar fasciae have traditionally been discounted owing to their small capacity for force generation or transmission. TrA's early onset of contraction enables it to act briefly and in relative isolation (together with MLF, PLF and deep multifidus) on the lumbar segments. Its small cross-sec­ tional area (Richardson et a11999) is well suited to generate small tensile loads and the fasciae are well structured to tol­ erate these. Fascial tensile forces are noted to increase with applied tension on TrA up to 10 N, but not proportionally above this (Barker et aI2004). T he low forces applied to fascia by Barker et al (2004a) and Tesh (1986) support the proposed capacity of fasciae to influence movement at all segments via an early onset, sub­ maximal contraction of TrA. Although Tesh's study (Tesh 1986) simulated only about 20% of the maximum (90 N) lat­ eral tension estimated by the author to be generated in the



PLF, this was sufficient to reduce neutral zone movement (see Fig. 11.5). The findings support the basis for exercises incorporating submaximal contraction of TrA (Richardson & Jull 1995, Richardson et al 1999) for management and prevention of low back pain. Of interest, Tesh himself did not consider the effects of fascial tension to be adequate to influence segmental move­ ment (Tesh 1986). In view of subsequent research recogniz­ ing the importance of limitation of the neutral zone (Panjabi 1992b) and the potential importance of an early onset of TrA contraction (Hodges & Richardson 1996), the effects, although relatively small, may still be significant.
P lanar stability

Attachments to transverse and spinous processes (Farfan 1975), along with obliquely oriented fibres, permit the MLF and PLF to resist movement in all three movement planes. Vertebral processes maximize their leverage on segmental movement, providing the fasciae with approximately 7 em moment arms (Tesh 1986) in the neutral position. Contraction of TrA prior to trunk perturbations has also been proposed to influence segmental neutral zone movement in multiple planes (Hodges & Richardson 1997b).
Coronal stability

on the convex side (Tesh et al 1987). The asymmetry was hypothesized to create a corrective moment that tended to approximate the transverse processes on that side (Fig. 11.6). It was therefore suggested that the MLF contributes increas­ ingly to stability during trunk lateral flexion, while the PLF was able to be cut without affecting pressure (Tesh 1986). The study indicated a tendency for tension from TrA and the MLF to return the spine to the neutral position (Tesh et a11987), but did so using pressures associated with a strong contraction of TrA (120 mmHg) and from end of range lat­ eral flexion rather than from the neutral position (in which fascial fibre angles would be symmetrical). Although indicative of a significant role for the MLF in the coronal plane, its application to segmental neutral zone motion is less clear than concurrent studies in the sagittal plane gen­ erating submaximal fascial tension (Tesh 1986) and requires further investigation. EMG studies do, however, indicate an onset of contraction of TrA prior to coronal trunk perturba­ tions, so both MLF and PLF are likely to affect (inner range) coronal plane stability to some extent. Fibre angles dictate that this tensile effect will be more substantial in the trans­ verse plane.
Sagittal stability

From their intra-abdominal balloon experiments, Tesh and colleagues (Tesh 1986, Tesh et a11987) proposed that if suffi­ cient tension was generated (e.g. by TrA contraction) during lateral flexion, the MLF could contribute to segmental coro­ nal plane stability by generating greater longitudinal tension

Since both spinous and transverse processes lie behind the instantaneous axis of rotation for flexion, the PLF and MLF may both contribute to sagittal plane stability, the PLF more so due to a longer moment arm. Tension on the PLF may be transmitted to vertebrae via the supraspinous and inter­ spinous ligaments, helping to limit sagittal rotation and shear movements respectively (Farfan 1995). Tesh's experiments (Tesh 1986) provide compelling evi­ dence to support the role of the PLF in limiting segmental neutral zone movement in the sagittal plane (see Fig. 11.5). Fascial tension increased inner range stiffness during test­ ing of all lumbar segments from four cadavers and the dif­ ference was reproducible if tension was detached then reapplied. The MLF has not been similarly tested, but has been shown to transmit lateral tension loads effectively (Barker et al 2004). EMG behaviour of TrA during sagittal trunk perturbations and with low back pain (Hodges & Richardson 1996) is consistent with such an (inner range) stabilizing role for both the MLF and PLF. Tensile transmis­ sion from TrA to the MLF may also help explain Kaigle et aI finding, in an in vivo porcine model, that removal of 's lumbar transverse processes (following facet joint injury) resulted in an increase in the neutral zone range for seg­ mental sagittal rotation (Kaigle et al 1995).
Transverse stability



The MLF i n l a tera l flexion. Asymmetry i n MLF fi b re

a n g les d u ri n g latera l flexi o n creates a co rrective moment. Ada pted from Tesh et al Key: MlF


m i d d l e l u m ba r fascia.

Very little axial rotation occurs in the lumbar spine owing to the orientation of facet joints (Bogduk 1997). Accurate testing of this movement is consequently difficult and no known studies have quantified the effects of fasciae on seg­ mental rotation. However, minimal movement may be required to produce injury in this plane (Gracovetsky & Farfan 1986, Hickey & Hukins 1980) and twisting is com-

The l umbar fasciae and segmental control

15 1

monly implicated in low back injury (Kelsey et al 1984, Marras et al 1993, 1995, Mundt et al 1993). Appropriately timed influences on the neutral zone may be particularly crucial for preventing injury in this plane. Although EMG behaviour of IrA has not been reported in response to limb movements in the transverse plane, movements in sagittal and coronal planes are likely to gen­ erate alternating rotary demands in the transverse plane (Hodges & Richardson 1997b). On this basis, one might pre­ dict, IrA also demonstrates an early onset of contraction prior to transverse trunk peturbations. Anatomical and transverse loading studies indicate IrA and the MLF and PLF are well structured to resist tension in the transverse plane (Barker et al 2004a, 2004b, Iesh 1986). Their effect on segmental movement might be expected, from fibre directions and the results of stiffness testing, to be several times the magnitude of those observed in the sagit­ tal plane, and effected on a smaller range of motion.
Fascial disruption

The lumbar fasciae are frequently disrupted by injury or surgery. The MLF may be torn when its attachments are avulsed (Marshall 2001) and the PLF is routinely cut during lumbar spine surgery, lumbar fasciectomy or taking bone grafts from the iliac crest. The contribution of the relevant fascia(e) to segmental control is then compromised and one might expect the resultant deficiency to increase the indi­ vidual's chance of low back injury. Io date, however, this has not been reported, with management of the above con­ ditions being largely symptomatic (Howarth & Petrie 1964). It has been suggested that reattachment of the PLF to midline structures following surgery (Crock & Crock 1988) or the use of horizontal incisions (Aspden 1992) during spinal surgery may help to minimize rehabilitation time and impairment of spinal stability. During iliac crest bone grafts, an incision through the midline rather than iliac attachments of the PLF, followed by reattachment, has been reported to reduce postoperative pain over the harvest site l (Hutchinson & DaI 1994). Further research into disruption of these mechanisms and the effects of surgical intervention is required.

loads from IrA to all lumbar vertebrae. Although the MLF may provide a more effective and isolated route, to date only the PLF has been demonstrated to limit segmental neutral zone movement (in the sagittal plane). EMG studies in healthy subjects indicate an early onset of contraction of IrA occurs prior to trunk perturbations in several direc­ tions, and under these circumstances both MLF and PLF are predicted to limit excess intersegmental movement occur­ ring in all planes. Disruption of the early contraction onset of IrA, as observed with low back pain, will eliminate this fascial influence on the neutral zone and is likely to increase pre­ disposition to injury. The MLF and PLF also provide a mechanism for continuous proprioceptive feedback from each lumbar segment, with disruption of innervation possi­ bly contributing to reduced segmental control in patients with chronic low back pain. In addition to its segmental roles, the PLF has more global effects during activities that recruit its attached or enclosed global muscles and these may contribute to com­ pression across the SIJ and lumbar spine as well as increase the effectiveness of paraspinal muscle contraction (respec­ tively). Such global roles are effected on an underlying requirement for restriction of segmental movement, influ­ enced by local muscle activity and transverse fascial tension from IrA. Further work elaborating the effects of MLF and PLF on segmental movement and the consequences of fas­ cial disruption may provide greater insight into their roles in segmental control.

anterior layer of lumbar fascia middle layer of lumbar fascia posterior layer of lumbar fascia lumbar fasciae muscle attachments local muscles global muscles biomechan ical roles longitudi nal tension generation

hydraulic amplifier effect lumbar segmental stability load transfer across the midline sacroiliac stability proprioception planar stability coronal stability sagittal stability transverse stability disruption segmental stability magnitude of forces

The MLF and PLF are particularly well structured for trans­ verse tension and capable of transmitting even small tensile

1995 Mechanical testing of the spine: an appraisal of 20(19): 2151-2156 Adams M A, Green T P 1993 Tensile properties of the annulus fibrosus.
Adams M A methodology, results, and conclusions. Spine Aspden R M

1992 Review of the functional anatomy of the spinal

5: 372-387

ligaments and the lumbar erector spinae muscles. Clinical Anatomy

Aspden R M, Bornstein N H, Hukins D W Journal of Anatomy

1987 Collagen organisation

I: The contribution of fibre-matrix interactions to tensile stiffness
and strength. European Spine Journal

2: 203-208

in the interspinous ligament and its relationship to tissue function.

155: 141-151



Barker P J, Briggs C A 1999 Attachments of the posterior layer of lumbar fascia. Spine 24(17): 1 757-1764 Barker P J, Briggs C A, Bogeski G 2004a Tensile transmission across the lumbar fasciae in unembalmed cadavers: effects of tension to various muscular attachments. Spine 29(2): 129-138 Barker P J, Urquhart D M, Briggs C A, Story I, Fahrer M 2004b The middle layer of lumbar fascia and muscle attachments to lumbar transverse processes: implications for fracture and segmental control. (submitted) Bednar D A, Orr F W, Simon G T 1995 Observations on the pathomorphology of the thoracolumbar fascia in chronic mechanical back pain: a microscopic study. Spine 20(10): 1161-1164 Bergmark A 1989 Stability of the lumbar spine: a study in mechanical engineering. Acta Orthopaedica Scandinavica 230 (Supp!.): 1-54 Bogduk N 1997 Clinical anatomy of the lumbar spine and sacrum. Churchill Livingstone, New York, pp 115-117 Bogduk N, Macintosh J E 1984 The applied anatomy of the thoracolumbar fascia. Spine 9(2): 1 64-170 Bogduk N, Johnson G, Spalding D 1998 The morphology and biomechanics of latissimus dorsi. Clinical Biomechanics (Bristol, Avon) 13(6): 377-385 Breathnach A S 1965 Frazer's Anatomy of the human skeleton. J. & A. Churchill, London, pp 35-37 Carman B J, Blanton P L, Biggs N L 1972 Electromyographic study of the anterolateral abdominal musculature utilizing indwelling electrodes. American Journal of Physical Medicine 51(3): 113-129 Carr D, Gilbertson L, Frymoyer J, Krag M, Pope M 1985 Lumbar paraspinal compartment syndrome: a case report with physiologic and anatomic studies. Spine 10(9): 816--820 Cholewicki J, McGill S M 1996 Mechanical stability of the in vivo lumbar spine: implications for injury and chronic low back pain. Clinical Biomechanics (Bristol, Avon) 11(1): 1-15 Cresswell A G, Grundstrom H, Thorstensson A 1992 Observations on intra-abdominal pressure and patterns of abdominal intra­ muscular activity in man. Acta Physiologic a Scandinavica 144(4): 409-418 Crock H V, Crock M C 1988 A technique for decompression of the lumbar spinal cana!' Neuro-Orthopedics 5: 96--99 Davies-Colley J N C 1894 The muscles. J & A Churchill, New York, pp 433-434 Davis J R, Mirka G A 2000 Transverse-contour modeling of trunk muscle-distributed forces and spinal loads during lifting and twisting. Spine 25(2): 1 80-189 Dijkstra P F, Vleeming A, Stoeckart R 1989 Complex motion tomography of the sacroiliac joint: an anatomical and roentgenological study. ROFO Fortschr Geb Rontgenstr Nuklearmed 150(6): 635-642 Donisch E W, Basmajian J V 1972 Electromyography of deep back muscles in man. American Journal of Anatomy 133(1): 25-36 Fairbank J C, O'Brien J P 1980 The abdominal cavity and thoraco­ lumbar fascia as stabilisers of the lumbar spine in patients with low back pain. In: Engineering aspects of the spine. Mechanical Engineering Publications for the Institution of Mechanical Engineers, London, pp 83-88 Farfan H F 1973 Mechanical disorders of the low back. Lea & Febiger, Philadelphia, pp 1-39, 171-197 Farfan H F 1975 Muscular mechanism of the lumbar spine and the position of power and efficiency. Orthopedic Clinics of North America 6(1): 135-144 Farfan H F 1995 Form and function of the musculoskeletal system as revealed by mathematical analysis of the lumbar spine: an essay. Spine 20(13): 1462-1474 Farfan H F, Cossette J W, Robertson G H, Wells R V, Kraus H 1970 The effects of torsion on the lumbar intervertebral joints: the role of torsion in the production of disc degeneration. Journal of Bone and Joint Surgery (American volume) 52(3): 468-497

Floyd W F, Silver P H S 1950 Electromyographic study of patterns of activity of the anterior abdominal wall muscles in man. Journal of Anatomy 84: 132-145 Floyd W F, Silver P H S 1955 The function of the erectores spinae muscles in certain movements and postures in man. Journal of Physiology 129: 184-203 Gallaudet B B 1931 A description of the planes of fascia of the human body. Columbia University Press, New York, pp 2-59 Gallios M, Japiot M 1925 Architecture interieure des vertebres (statique et physiologie de la colonne vertebrale). Revue de Chirurgie 63: 687-708 Gracovetsky S, Farfan H 1986 The optimum spine. Spine 11(6): 543-573 Gracovetsky S, Farfan H F, Lamy C 1977 A mathematical model of the lumbar spine using an optimized system to control muscles and ligaments. Orthopedic Clinics of North America 8(1): 135-153 Gracovetsky S, Farfan H F, Lamy C 1981 The mechanism of the lumbar spine. Spine 6(3): 249-262 Gracovetsky S, Farfan H, Helleur C 1985 The abdominal mechanism. Spine 10(4): 317-324 Hickey D S, Hukins D W 1980 Relation between the structure of the annulus fibrosus and the function and failure of the intervertebral disc. Spine 5(2): 106--116 Hides J A, Jull G A, Richardson C A 2001 Long-term effects of specific stabilizing exercises for first-episode low back pain. Spine 26(11): E243-248 Hodges P W, Richardson C A 1996 Inefficient muscular stabilization of the lumbar spine associated with low back pain: a motor control evaluation of transversus abdorninis. Spine 21 (22): 2640-2650 Hodges P W, Richardson C A 1997a Relationship between limb movement speed and associated contraction of the trunk muscles. Ergonomics 40(11): 1220-1230 Hodges P W, Richardson C A 1997b Feedforward contraction of transversus abdorninis is not influenced by the direction of arm movement. Experimental Brain Research 114(2): 362-370 Hodges P W, Richardson C A 1997c Contraction of the abdominal muscles associated with movement of the lower limb. PhYSical Therapy 77(2): 132-142 Hodges P W, Richardson C A 1998 Delayed postural contraction of transversus abdorninis in low back pain associated with movement of the lower limb. Journal of Spinal Disorders 11(1): 46--56 Hodges P W, Cresswell A, Thorstensson A 1999 Preparatory trunk motion accompanies rapid upper limb movement. Experimental Brain Research 124(1): 69-79 Hodges P W, Moseley G L, Gabrielsson A, Gandevia S C 2001 Experimentally induced low back pain causes changes in motor control of the trunk muscles. Fourth Interdisciplinary World Congress on Low Back and Pelvic Pain, Montreal, European Conference Organizers, pp 184-185 Howarth M B, Petrie J G 1964 Injuries of the spine. Williams & Wilkins, Baltimore, p 265 Hukins D W L 1984 Connective tissue matrix: topics in molecular and structural biology. Macmillan, London, pp 211-240 Hukins D W L 1985 Composition and properties of connective tissues. Trends in Biochemical Science 10: 260-264 Hukins D W L, Aspden R M, Hickey D S 1990a Thoracolumbar fascia can increase the efficiency of the erector spinae muscles. Clinical Biomechanics 5: 30-34 Hukins D W L, Kirby M C, Sikoryn T A, Aspden R M, Cox A J 1990b Comparison of structure, mechanical properties, and ftmctions of lumbar spinal ligaments. Spine 15(8): 787-795 Hutchinson M R, Dall B E 1994 Midline fascial splitting approach to the iliac crest for bone graft: a new approach. Spine 19(1): 62-{)6 Jull G A, Richardson C A 2000 Motor control problems in patients with spinal pain: a new direction for therapeutic exercise. Journal of Manipulative and Physiological Therapeutics 23(2): 115-117 Kaigle A M, Holm S H, Hansson T H 1995 Experimental instability in the lumbar spine. Spine 20: 421-430

The l umbar fasciae and segmental control


Kelsey J L, Githens P B, White A A 3rd et al 1984 An epidemiologic study of lifting and twisting on the job and risk for acute prolapsed lumbar intervertebral disc. Journal of Orthopedic Research 2(1): 61-66 McGill S M, Norman R W 1988 Potential of lumbodorsal fascia forces to generate back extension moments during squat lifts. Journal of Biomedical Engineering 10(4): 312-318 McGill S M, Norman R W 1993 Low back biomechanics in industry: the prevention of injury through safer lifting. In: Grabiner M D (ed) Current issues in biomechanics. Human Kinetics Publishers, Champaign, illinois, pp 69-121 Macintosh J E, Valencia F V, Bogduk N, Munro R R 1986 The morphology of the human lumbar multifidus. Clinical Biomechanics 1(4): 196-204 Marras W S, Lavender S A, Leurgans S E et al 1993 The role of dynamic three-dimensional trunk motion in occupationally-related low back disorders: the effects of workplace factors, trunk position, and trunk motion characteristics on risk of injury. Spine 18(5): 617-628 Marras W S, Lavender S A, Leurgans S E et al 1995 Biomechanical risk factors for occupationally related low back disorders. Ergonomics 38(2): 377-410 Marshall R 2001 Living anatomy: structure as the mirror of function. Melbourne University Press, Melbourne, Australia, pp 280-283 Mens J M, Vleeming A, Stoeckart R, Starn H J, Snijders C J 1996 Understanding peripartum pelvic pain: implications of a patient survey. SFine 21(11): 1363-1369; discussion 1369-1370 Mens J M, Snijders C J, Starn H J 2000 Diagonal trunk muscle exercises in peripartum pelvic pain: a randomized clinical trial. Physical Therapy 80(12): 1164-1173 Minns R J, Soden P D, Jackson D S 1973 The role of the fibrous components and ground substance in the mechanical properties of biological tissues: a preliminary investigation. Journal of Biomechanics 6(2): 153-165 Mirka G, Kelaher D, Baker A, Harrison A, Davis J 1997 Selective activation of the external oblique musculature during axial torque production. Clinical Biomechanics (Bristol, Avon) 12(3): 172-180 Mooney V, Pozos R, Vleerning A, Gulick J, Swenski D 2001 Exercise treatment for sacroiliac pain. Orthopedics 24(1): 29-32 Moseley G L, Hodges P W, Gandevia S C 2002 Deep and superficial fibers of the lumbar multifidus muscle are differentially active during voluntary arm movements. Spine 27(2): E29-36 Mundt D J, Kelsey J L, Golden A L et al 1993 An epidemiologic study of non-occupational lifting as a risk factor for herniated lumbar intervertebral disc. Northeast Collaborative Group on Low Back Pain. Spine 18(5): 595-602 O'Sullivan P, Twomey L, Allison G, Sinclair J, Miller K 1997a Altered patterns of abdominal muscle activation in patients with chronic low back pain. Australian Journal of Physiotherapy 43(2): 91-98 O'Sullivan P B, Twomey L T, Allison G T 1997b Evaluation of specific stabilizing exercise in the treatment of chronic low back pain with radiologic diagnosis of spondylolysis or spondylolisthesis. Spine 22(24): 2959-2967 Panjabi M M 1992a The stabilizing system of the spine. I: Function, dysfunction, adaptation, and enhancement. Journal of Spinal Disorders 5(4): 383-389 Panjabi M M 1992b The stabilizing system of the spine. II: Neutral zone and instability hypothesis. Journal of Spinal Disorders 5(4): 390-396

Panjabi M, Aburni K, Duranceau J, Oxland T 1989 Spinal stability and intersegmental muscle forces: a biomechanical model. Spine 14(2): 194-200 Peck D, Nicholls P J, Beard C, Allen J R 1986 Are there compartment syndromes in some patients with idiopathic back pain? Spine 11(5): 468-475 Poirier P 1901 Myologie. In: Poirer P, Charpy A (eds) Traite d'anatomie humaine. Masson et Compagnie, Paris, vol 2, pt 1, P 497 Richardson C A, Jull G A 1995 Muscle control-pain control. What exercises would you prescribe? Manual Therapy 1(1): 2-10 Richardson C A, Jull G A, Hodges P W, Hides J A 1999 Therapeutic exercise for spinal segmental stabilization in low back pain. Harcourt Brace, London, pp 4-59 Richardson C A, Snijders C J, Hides J A, Damen L, Pas M S, Storm J 2002 The relation between the transversus abdominis muscles, sacroiliac joint mechanics, and low back pain. Spine 27(4): 399-405 Sharpey W, Thomson A, Cleland J (eds) 1867 Quain's Elements of anatomy. James Walton Publishing, London, pp 248-253 Styf J, Lysell E 1987 Chronic compartment syndrome in the erector spinae muscle. Spine 12(7): 680-682 Tesh K M 1986 The abdominal muscles and vertebral stability. PhD thesis, Bioengineering Unit, University of Strathclyde, pp 166-349 Tesh K M, Dunn J S, Evans J H 1987 The abdominal muscles and vertebral stability. Spine 12(5): 501-508 Testut L, Latarjet A 1948 Traite d'anatomie humaine. G. Doin & Compagnie, Paris, pp 944-952 Urquhart D M, Barker P J, Hodges P W, Story I, Briggs C A 2004 Regional morphology of tranversus abdominis, obliquus intemus and obliquus externus abdominis. (submitted) Vleeming A, Stoeckart R, Snijders C J 1989 The sacrotuberous ligament: a conceptual approach to its dynamic role in stabilizing the sacroiliac joint. Clinical Biomechanics 4(4): 201-203 Vleeming A, Buyruk H M, Stoeckart R, Karamursel S, Snijders C J 1992 An integrated therapy for peripartum pelvic instability: a study of the biomechanical effects of pelvic belts. American Journal of Obstetrics and Gynecology 166(4): 1243-1247 Vleeming A, Pool-Goudzwaard A L, Stoeckart R, van Wingerden J P, Snijders C J 1995 The posterior layer of the thoracolumbar fascia: its function in load transfer from spine to legs. Spine 20(7): 753-758 Vleeming A, Pool-Goudzwaard A L, Hammudoghlu D, Stoeckart R, Snijders C J, Mens J M 1996 The function of the long dorsal sacroiliac ligament: its implication for understanding low back pain. Spine 21(5): 556-562 Williams P W, Bannister L H, Berry M M et al (eds) 1995 Gray's Anatomy. Churchill Livingstone, New York, pp 809-829 Wingerden J P V, Vleerning A, Buyruk H M, Raissadat K 2001 Muscular contribution to force closure: sacroiliac joint stabilisation in vivo. Fourth Interdisciplinary World Congress on Low Back and Pelvic Pain, Montreal, European Conference Organizers, pp 153-159 Yahia L, Rhalmi S, Newman N, Isler M 1992 Sensory innervation of human thoracolumbar fascia: an immunohistochemical study. Acta Orthopaedica Scandinavica 63(2): 195-197 Yahia L H, Pigeon P, DesRosiers E A 1993 Viscoelastic properties of the human lumbodorsal fascia. Journal of Biomedical Engineering 15(5): 425-429





Neurophysiology of pain and pain modulation
A. Wright, M. Zusman

155 155 156 156 156 157

Hyperalgesia and allodynia

Peripheral regulatory mechanisms 'Inflammatory soup'
156 157

Nocic!!ptor activation and sensitization Multiple forms of peripheral sensitization Membrane ion channels G-protein coupled intracellular cascades Inhibition of hyperpolarization Indirect mechanisms Trophic factors
157 158 158 158 158 158 159 157 157

Silent nociceptors Phenotype change

Spinal cord regulatory mechanisms Wide dynamic range cells Nociceptive-specific cells Central sensitization
159 160 160 161

NMDA receptor activation Nitric oxide production Long-term potentiation Differential sensitization Trophic factors

Neuroanatomical reorganization Functional brain imaging Attentional processes Emotion
163 162 162

161 162

Central integration of nociceptive input

In clinical practice manual therapists often see the impact of pain on an individual. Pain can change rapidly from no pain, or minimal levels of pain, to a situation where the pain experience is so severe and pervasive that it drives all of the person's behaviour. Whiplash injury, acute back injury, major fracture or other acute trauma can all provide an indication of the impact of pain on previously pain-free individuals. The propensity for pain and disability to per­ sist in the absence of obvious ongoing primary peripheral pathology is both baffling and challenging. This chapter includes evidence for peripheral and central nervous sys­ tem mechanisms that might contribute to the enhancement and maintenance of clinically observed symptoms and signs. It also describes forebrain mechanisms that could allow emotions and cognitions to effectively sensitize spinal cord pain pathway neurons, possibly providing a mechanism for psychosocial factors to enhance pain per­ ception. Potential interactions between the nociceptive and motor systems are considered. The level of change that occurs in the behaviour of individuals who experience pain implies a very marked enhancement of nociceptive system function and, consequent upon this, enormous neuroplas­ ticity and change in many aspects of central nervous system function. Mechanisms contributing to that neuroplasticity will be described.

Forebrain mediated modulatory systems Loss of inhibition Facilitation
166 166 166 165


Somatomotor dysfunction Vicious cycle model Emerging models Conclusion
168 167

Enhanced withdrawal reflexes Pain adaptation model
167 167

A large body of research has developed describing the ways in which nociceptive system function can be enhanced and pointing to the effects of this on somatomotor function. The nociceptive system is normally a quiescent system requir­ ing strong, intense, potentially damaging stimulation before it becomes activated. Yet once an individual is expe­ riencing pain, relatively innocuous stimuli activate the sys­ tem and trigger pain perception. This altered perceptual state is encompassed by the phenomena of hyperalgesia, an exaggerated or increased response to a noxious stimulus,



and allodynia, the production of pain by a stimulus that would not normally be painful (Merskey & Bogduk 1994). It is now clear that after injury there may be distinct dif­ ferences between the mechanisms leading to hyperalgesia in injured and uninjured tissue. It is also apparent that the time course and extent of hyperalgesia is different for dif­ ferent forms of injury. It is widely recognized that prolonged C fibre input evokes two distinct forms of cutaneous hyperalgesia; these have been termed primary and secondary hyperalgesia (Hardy et al 1950). It is suggested that primary hyperalge­ sia is predominantly due to peripheral sensitization of noci­ ceptors whereas secondary hyperalgesia is dependent on the process of central sensitization in cells processing noci­ ceptive information at the spinal cord level (Torebjork et al 1992).

ferent stimulation modalities (Mizamura & Kumazawa 1996). It is therefore important to note that nociceptors may become differentially sensitized to thermal, mechan­ ical and chemical stimuli. An individual nociceptor can potentially exhibit sensitization to thermal stimuli, for example, while retaining normal sensitivity to mechanical or chemical stimuli.
'Inflammatory soup'

There is now a large body of research investigating modu­ latory mechanisms in both the periphery and the central nervous system. It has become apparent that these processes are of great importance in altering nociceptive system function following injury.
Nociceptor activation and sensitization

Many early studies pointed to sensitization of peripheral nociceptors as a mechanism underlying the increased sen­ sitivity to subsequent stimulation that takes place following tissue injury. It is apparent that many peripheral nocicep­ tors are polymodal, in the sense that they respond to chem­ ical as well as mechanical and thermal nociceptive stimulation (Kumazawa 1996). It is also apparent that chemical mediators released into the tissues because of tis­ sue injury promote sensitization of peripheral nociceptors. Key mediators which have been identified include bradykinin, serotonin, histamine, potassium ions, adeno­ sine triphosphate, protons, prostaglandins, nitric oxide, leukotrienes, cytokines and growth factors (Dray 1995). The effects of these mediators involve binding to specific recep­ tors, activation of ion channels for depolarization, activa­ tion of intracellular second messenger systems, release of a range of neuropeptides to promote neurogenic inflamma­ tion and alteration of neuronal properties by modifying gene transcription (Bevan 1996, Dray 1995). Release of a range of inflammatory mediators triggers phosphorylation and activation of a number of receptors and second mes­ senger systems (Mizamura & Kumazawa 1996). The actions of chemical mediators normally fall into one of two categories: either direct activation of nocicep­ tive afferents or sensitization so that subsequent stimula­ tion leads to an enhanced response. While polymodal receptors respond to a range of stimuli, it is apparent that different molecular receptors and second messenger sys­ tems are involved in excitation and sensitization for dif-

One of the most fundamental influences on nociceptor sensitivity is the pH of the surrounding tissue. High local proton concentrations are known to occur in many inflam­ matory states and the consequent reduction in pH can contribute to sensitization and activation of polymodal nociceptors (Handwerker & Reeh 1991, 1992, Reeh & Steen 1996). Altered pH of the local chemical environment of peripheral nociceptors is a particularly important factor in inducing mechanical sensitization and ischaemic pain (Dray 1995, Steen et al 1992). Combinations of inflamma­ tory mediators and combination of chemical mediators with altered tissue pH appear to be more effective in induc­ ing sensitization than individual chemical mediators alone (Handwerker & Reeh 1991 ). Thus in the natural situation it appears to be a blend of chemical mediators, termed 'inflammatory soup' by Handwerker & Reeh, that produces sensitization of peripheral nociceptors (Handwerker & Reeh 1991).
Multiple forms of peripheral sensitization

Endogenous chemicals act on a variety of receptors, acti­ vate three major intracellular second messenger systems and influence different ion channels (Dray 1995, Mizamura & Kumazawa 1996), resulting in distinctions between ther­ mal, mechanical and chemical sensitization in specific pop­ ulations of nociceptors (Mizamura & Kumazawa 1996). For example, prostaglandins may induce sensitization to chemical mediators at much lower concentrations than those required to induce sensitization to heat stimuli (Mizamura & Kumazawa 1996). The vanilloid receptor VR1 has been identified as a specific molecular mechanism for thermal hyperalgesia, as well as sensitization following capsaicin administration (Cesare et al 1999). The receptor will normally respond to temperatures in excess of 45'C. However, under low pH conditions there is a significant reduction in the activation threshold of the receptor such that VR1 may be activated at normal tissue temperatures (Harding 1999, Thacker 2002). This could potentially con­ tribute to ongoing pain. As noted above, nociceptor activation and sensitization can be produced through a number of different mecha­ nisms. It can occur as a result of a direct influence of chem­ ical mediators on membrane ion channels. It can also occur as a result of the action of chemical mediators on G-proteins and second messenger systems.

Neurophys i o l ogy of p a i n a n d p a i n m odu l a tion


Membrane ion channels

Indirect mechanisms

One e�ample is the fact that increased proton concentration results in increased membrane permeability to cations and sustained changes in neuronal activation. The mechanism of action of protons appears to be very similar to that of exogenously applied capsaicin (Dray 1995). Proton binding leads to phosphorylation of sodium channels resulting in a more sustained open state of the receptor and greater sodium influx to the cell (Kingsley 2000). Adenosine triphosphate, bradykinin, serotonin and prostaglandins may act on receptors that produce changes in potassium ion permeability (Dray 1995). The overall consequence is a sig­ nificantly increased number of action potentials generated by peripheral nociceptors.
G-protein coupled intracellular cascades

Sensitization following the release of cytokines and leukotrienes appears to occur via indirect mechanisms whereby these agents stimulate other cells to release sen­ sitizing agents. For example, leukotriene B4 stimulates the release of 8R,15SdiHETE from leucocytes, and this then acts to sensitize polymodal nociceptors (Levine et al 1993). Some of these agents may also act to induce recep­ tors for other inflammatory mediators (Rang & Urban 1995). In addition, Ca++ and calmodulin can activate nitric oxide synthase to trigger the production of nitric oxide. Nitric oxide functions as a messenger between neurons and surrounding tissues. As it diffuses widely through the tis­ sues it can induce relaxation of vascular smooth muscle and it may contribute to the spread of sensitization in the peripheral tissues (Anbar and Gratt 1997).
Trophic factors

One of the most effective mechanisms for sensitizing noci­ ceptors is activation of G-protein coupled second messen­ ger systems. These intracellular cascades result in very significant amplification of the neuronal signal. Binding of a chemical mediator to a G-protein receptor results in acti­ vation of that receptor and guanosine triphosphate (GTP) binding. The activated G-protein is released into the cytosol and may then bind to an appropriate enzyme such as adenylate cyclase causing it to catalyse a second messenger protein. In this case the second messenger will be cyclic adenosine monophosphate (cAMP), which in turn will acti­ vate protein kinase A (PKA). PKA is then responsible for phosphorylation of membrane ion channels and increased ion flux. Because binding to one G-protein receptor can lead to activation of not one but many ion channels, this process has the ability to greatly amplify the initial signal. It is also apparent that kinins such as bradykinin may act on appropriate receptors causing phospholipase C acti­ vation among other effects. This leads to the release of intracellular calcium and diacylglycerol (DAG), which in turn activates protein kinase C (PKC). These pathways lead to activation of ion channels to increase membrane permeability, particularly for sodium and calcium ions (Dray 1995). Increased intracellular calcium ion concentra­ tion also leads to the release of neuropeptides such as sub­ stance P and stimulation of arachidonic acid production (Dray 1995).
Inhibition of hyperpolarization

Another mechanism by which peripheral sensitization can occur is inhibition of the hyperpolarization that occurs after impulse generation. This slow after-hyperpolarization lim­ its the number of action potentials that can be generated following stimulation. Prostaglandins and bradykinin act to inhibit this phenomenon, allowing the neuron to fire repetitively (Dray 1995). This may also be one of the mech­ anisms activated by serotonin (Dray 1996).

There is increasing evidence for the role of nerve growth factor (NGF) as a mediator of hyperalgesia (Anand 1995). Its actions include triggering mast cell degranulation, stim­ ulating the release of neuropeptides and regulating other proteins such as proton activated ion channels (Anand 1995, Dray 1995, 1996, Shu & Mendell 1999). The induced hyperalgesia may be reduced by the administration of anti­ nerve growth factor antibodies (Woolf et al 1994). It has been suggested that NGF may be particularly important for thermal sensitization and that it may condition the response of the VR1 receptor to capsaicin or thermal stimu­ lation (Shu & Mendell 1999). Mechanical hyperalgesia appears to be induced over a much longer time period (Shu & Mendell 1999). Blockade of nerve growth factor function by the administration of the nerve growth factor specific tyrosine kinase receptor A coupled to human immunoglob­ ulin-y (trkA-IgG) prevents the development of thermal and mechanical hyperalgesia following joint inflammation (McMahon et al 1995). The trkA-IgG fusion molecule has the effect of binding nerve growth factor and reducing the amount of free NGF present in the tissues. Other neurotrophins include brain-derived neurotro­ phic factor (BDNF) and glial cell line-derived neu­ rotrophic factor (GDNF) . Although expressed by peripheral afferents their primary effect appears to be related to modulation of central nervous system neurons. Activity induced phosphorylation of cAMP leads to acti­ vation of a gene transcription factor known as cyclic adenosine monophosphate response element binding pro­ tein (Woolf & Costigan 1999). This mediates transcription of BDNF within the dorsal root ganglion, which is then axoplasmically transported to presynaptic terminals in the dorsal horn. Subsequent peripheral stimulation will result in significantly increased release contributing to central sensitiza tion.

1 58


The processes of peripheral sensitization are clearly one way in which nociceptive system activity can be increased in response to tissue injury. It is apparent that the sensiti­ zation process is relatively complex and that different forms of sensitization may develop depending on the nature of the injury or disease. Spontaneous discharge, reduced activation thresholds, inhibition of slow after­ hyperpolarization and increased discharge rates in response to suprathreshold stimulation contribute to increasing the nociceptive afferent input to the central nervous system. These mechanisms are of considerable importance in the immediate period after tissue injury. Where pain persists beyond the time required for tissue healing it is likely that they will make a lesser contribution to ongoing pain and hyperalgesia.
Silent nociceptors

It is now well established that in some tissues there is a sig­ nificant population of nociceptors that remain essentially inactive under normal conditions. These sleeping or silent nociceptors are activated because of tissue injury with con­ sequent release of chemical mediators and increased tissue hypoxia (Schmidt 1996). Schmidt estimates that they may represent approximately one third of the total nociceptor population in joints (Schmidt 1996). They appear to be pres­ ent in skin, joints and visceral tissue. At least 50% of vis­ ceral afferents may fall into this category (Mayer & Gebhart 1994). Once activated, these silent nociceptors exhibit marked sensitization with increased spontaneous discharge rates, reduced thresholds for evoked discharge and increased discharge rates in response to stimulation.
Phenotype change

1999). They may also contribute to the release of peptides and other chemical mediators responsible for peripheral sensitization and neurogenic inflammation. This conver­ sion of myelinated AP afferents with low activation thresh­ olds appears to be an additional mechanism for enhancing peripheral nociceptive input. Activation of nociceptors, sensitization of currently responsive nociceptors, recruitment of mechanically insen­ sitive or silent nociceptors and phenotype conversion of non-nociceptive afferents represent major mechanisms whereby tissue injury and inflammation can trigger both temporal and spatial summation of nociceptive afferent inputs to the central nervous system. Acting in concert, these mechanisms can contribute to substantial changes in peripheral nociceptive system function. Ultimately increased impulse activity in nociceptive neurons may be interpreted as pain at higher levels within the central nerv­ ous system. It is apparent, however, that there is not a con­ stant link between the degree of tissue damage and the level of pain experienced. Modulatory influences will be discussed later in this chapter.

The existence of both wide dynamic range and nociceptive­ specific neurons within the dorsal hom of the spinal cord is now widely recognized. Both cell types contribute to changes in nociceptive system function following injury (Fig. 12.1).
Wide dynamic range cells

A further mechanism contributing to enhanced activity of the nociceptive system is phenotype transformation (Woolf & Costigan 1999). It was proposed that in the situation where inflammation has been present for some time, tran­ scriptional changes in gene expression might result in phe­ notypic changes in some myelinated AP afferents such that these fibres acquire the neurochemical properties of unmyelinated C fibres. Normally only C fibre (and a few AO) nociceptors contain the peptides necessary for clini­ cally relevant central sensitization (e.g. substance P and cal­ citonin gene related peptide, CGRP) (Ma & Woolf 1995, 1996). However, it appears that the combination of inflam­ mation, mechanical stimulation and release of growth fac­ tor molecules such as NGF bring about a genetically mediated phenotypic 'switch' (Woolf & Costigan 1999). Under these circumstances, myelinated afferents begin to express and release neuropeptides that are important for inducing long-term changes in neuronal sensitivity (Neumann et al 1996). Importantly, this means that these fibres can contribute to nociception by inducing sensitiza­ tion in central nervous system neurons (Woolf & Costigan

Wide dynamic range cells are particularly prevalent in the deeper laminae of the dorsal hom. They receive mput from both nociceptive and non-nociceptive afferent neurons and exhibit a graded response pattern related to the intensity of the afferent stimulus. If these neurons become sensitized and hyper-responsive they may discharge at a high rate fol­ lowing previously non-noxious stimulation (such as mild thermal or tactile stimulation) (Siddall & Cousins 1998). If the activity of the wide dynamic range neuron exceeds a threshold then the previously non-noxious stimulus will be perceived as painful. This phenomenon may, in part, pro­ vide the neurophysiological basis for the phenomenon of secondary hyperalgesia in which pain is perceived follow­ ing stimulation of normal uninjured tissue.
Nociceptive-specific cells

Nociceptive-specific cells are located predominantly in the superficial laminae of the dorsal hom, where they receive inputs from unmyelinated C fibre afferents. Under normal conditions, their response characteristics include a lack of impulse generation in response to non-noxious stimulation and a relatively sluggish response to intense noxious stim­ ulation of their peripheral receptive fields. However, fol-

N e u rophysiol ogy of p a i n a nd p a i n modu l ation

1 59






+ +

'0 C


8 g: :u a.

0 0 60 120


Q) .!!1 0

5.120 E


Squeeze 60

+ + +


60 Time/s


Figure 12.1

Response profiles of A: wide-dynamic range and


nociceptive-specific neurons in the spinothalamic tract. The receptive fields of the cells are indicated on the figures to the left. Excitatory receptive fields are indicated by + signs and inhibitory receptive fields are indicated by - signs. Reproduced from Willis,


lowing stimulation of peripheral nociceptive afferents and the development of central sensitization, the response char­ acteristics of nociceptive-specific neurons change (Cook et aI 1987). Subliminal inputs from myelinated afferent neu­ rons are enhanced and the cells begin to exhibit response characteristics that are similar to those of wide dynamic range neurons. Despite the fact that these cells can now be activated by non-noxious afferent inputs, it is likely that impulse activity generated by these neurons will contribute to pain perception at higher levels in the central nervous system. This may constitute another mechanism whereby normally non-nociceptive inputs from injured or uninjured tissues can contribute to pain perception. The processes of central sensitization influence both wide dynamic range and nociceptive-specific cells in the dorsal hom of the spinal cord. Following tissue injury, the response characteristics of these cells change such that nor­ mally non-nociceptive afferent inputs via myelinated affer­ ents can generate impulse activity that is likely to trigger pain perception.
Central sensitization

thinly myelinated AS (group III) primary afferents results in lasting increases in the excitability and responsiveness of both wide dynamic range and nociceptive-specific neurons in spinal cord pain pathways (Woolf & Salter 2000). The process of central sensitization (Woolf 1994) is an important aspect of neuroplasticity that contributes to enhanced acti­ vation of the nociceptive system in response to injury. This process may provide a link between the presence of pain and sensorimotor dysfunction in patients experiencing pain. Central sensitization describes the changes occurring at a cellular level to support the process of neuronal plas­ ticity that occurs in nociceptive system neurons in spinal cord and supraspinal centres, because of activation of the nociceptive system (Woolf 1994). The key to understanding the relevance of central sensi­ tization to the production and interpretation of clinically observed symptoms and signs is to appreciate that spinal cord pain pathway neurons make more than routine hard­ wired connections with their prescribed receptive fields. As noted above, they also have a vast number of redundant, 'subliminal' connections with surrounding areas and struc­ tures. Under normal circumstances, the redundant inputs are too weak to be effective. Hence, normally the appropri­ ate somatotopic map and sensory specificity are main­ tained. However, should the central neurons' usually high thresholds for excitation be significantly lowered for any reason, the potential exists for new connections to emerge. Under these circumstances, central pain pathway neurons display increased discharges to their normally effective inputs and discharges to previously ineffective inputs. As a consequence they respond to near and distant inputs that were formerly subliminal, normally inappropriate (e.g. along AP afferents) and non-somatotopic (Woolf & Doubell 1994). This contributes to some of the hallmarks of central sensitization in terms of expanded receptive fields and an exaggerated response to subsequent inputs. This is the situation that prevails to varying degrees soon after dorsal hom pain pathway neurons receive a barrage of input along unmyelinated and thinly myelinated periph­ eral nociceptive afferents (Woolf & Doubell 1994). As a result, in addition to the production of excessive pain from local pathological tissue (peripheral sensitization), clinical manifestations of central sensitization include widespread spontaneously arising or non-noxiously provoked pain in distant normal tissue, mechanical allodynia, referred pain and referred tenderness. This process is initiated by activity in peripheral nociceptors, particularly those associated with unmyelinated afferent neurons but it appears that the process can also be sustained in the absence of peripheral nociceptor input (Coderre & Melzack 1987, Woolf 1983).
NMDA receptor activation

There is now considerable evidence that, as well as produc­ ing changes at their peripheral terminals, more than tran­ sient stimulation of unmyelinated C fibre (group IV) and

Activity in peripheral nociceptors releases the excitatory amino acid neurotransmitter glutamate. Glutamate then binds to two major ion channel receptor populations,

1 60


a-amino-3-hydroxy1-5-meth y1-isoxazoleproprionic acid (AMPA) and N-methyl-D-aspartate (NMDA). These recep­ tors influence resting membrane potential and depolariza­ tion of dorsal horn nociceptive neurons. The NMDA receptor subtype in particular has been strongly implicated in the generation of central sensitization (Dickenson 1995, Mao et al 1995, Woolf 1994). Release of excitatory amino acids such as glutamate and concomitant release of excita­ tory neuropeptides such as substance P and neurokinin A from the presynaptic terminals of nociceptive afferents ini­ tiates a cascade of changes in postsynaptic spinal cord neu­ rons (Duggan et al 1988, 1990 Wilcox 1991). These include ; activation of G-protein linked metabotrophic receptors such as the metabotrophic glutamate receptor (mGluR) and the neurokinin 1 receptor (NK1). G-protein mediated acti­ vation of phospholipase C leads to the release of Ca++ from intracellular compartments. G-protein binding also leads to activation of protein kinase C, which in turn modulates ion channel activity (Mao et al 1995, Woolf 1994). These changes up-regulate NMDA receptors and enhance the neuron's responsiveness to subsequent excitatory amino acid release (Woolf 1994). One outcome of this alteration in NMDA receptor function is an increased Ca++ influx into the cell. Increased intracellular Ca++ concentration reduces transmembrane potential and activates a number of enzymes such as PKA, tyrosine kinases, PKC and calcium calmodulin kinase. These kinases phosphorylate a range of receptors, ion channels and gene transcription factors, all of which contribute to relatively long lasting changes in the excitability of the dorsal horn neurons. These processes are summarized in Figure 12.2.
Nitric oxide production


NMDA·R O ' M 2+ -==''-- -9 DAG IP3

tv II •

Y t--c

� �

Ca 2;C�Gene.expression Arg




Presynaptic effects?
Release of glutamate from primary afferent fibres triggers many changes within spinal cord neurons. Activation of the NMDA receptor induces an influx of Ca++ through Ca++ channels, while activation of metabotrophic glutamate receptors mobilizes Ca++ from intracellular compartments. Reproduced from Mao et al 1995. Key: PAF primary afferent fibre; Glu glutamate; AMPA /KA-R RS-a-aminO-3-hydroxy-5-methylisoxazole-4-propionic acid / kainic acid receptor; NMDA-R N-methyl-D-aspartate receptor; mGluR metabotrophic glutamate receptor; G guanosine triphosphate (GTP l binding protein; PLC phospholipase C; Ca++­ CM calcium-calmodulin complex; NOS nitric oxide synthase; L-Arg L-arginine; L-Cit L-citrulline; cGMP cyclic guanosine monophosphate; PKs protein kinases.
= = = = = = = = = = = = =

Figure 12.2

One other effect of increased intracellular Ca++ concentra­ tion is to trigger the production of nitric oxide, which has important second messenger functions within the cell and is thought to be capable of diffusing out of the cell to bring about increased activation of and release of neurotransmit­ ter from the primary afferent neuron (Meller & Gebhart 1993). Synthesis of nitric oxide is catalysed by the enzyme nitric oxide synthase, which is activated by binding of Ca++ / calmodulin complexes (Gordh et al 1995). Nitric oxide in turn is thought to activate guanylate cyclase, triggering an intracellular cascade that eventually leads to release of stored intracellular Ca++. The capacity of nitric oxide to dif­ fuse and influence adjacent neurons may be an important factor in the spread of sensitization that appears to occur in spinal cord neurons.
Long-term potentiation

Long-term potentiation (LTP) may be described as the increase in efficacy that occurs at a synapse following suit­ able prior activity. In other words, once primed by a bar­ rage of nociceptive impulses, subsequent stimuli, including

those that are normally ineffective, elicit a much greater response from the postsynaptic neuron. At the first synapse in the dorsal horn this increase in excitability is expressed as excitatory postsynaptic action potentials that then travel to the brain. LTP has been strongly linked with learning and memory at recognized centres in the brain, such as the hlppocampus and cerebellum (Bear et al 2001). The physiological and structural changes necessary for learning and memory con­ solidation have profound implications for the production and maintenance of clinical pain. Evidence linking LTP and central sensitization is quite compelling. For instance, research to date has shown that the type and parameters of nociceptive afferent stimuli sufficient to induce LTP are similar to those that cause central sensitization and hyper­ algesia. Moreover, demonstrable dorsal horn LTP and tis­ sue insult-induced hyperalgesia possess the same signal transduction pathways, time course and pharmacological

N e u rophysi ology of p a i n a n d p a i n m od u l ation


profile (Sandkiihler 2000). This suggests that LTP at Ao and C fibre synapses provides an attractive cellular model for injury-induced hyperalgesia (Sandkiihler 2000). The molecular changes currently thought to be responsi­ ble for expression of LTP are strongly linked to alterations in function of the AMPA receptor. However, the NMDA receptor, Ca++ influx and the activation of protein kinases are all involved since interruption of any one or all of these blocks the development of LTP. Nevertheless, at least two distinct changes that this cascade is capable of evoking in the AMPA receptor are of major importance. One is increased ionic conductance following AMPA receptor phosphorylation, the other is the insertion of entirely new AMPA receptors into the postsynaptic membrane. Notably, both of these events are mediated by protein kinases (PKC and calmodulin kinase, CaMK). In the case of the former, synaptic efficacy is significantly increased by enhanced single channel conductance due to increased open channel time. This is the result of protein kinase mediated phosphorylation (Lin et al 2002) . Acquisition by a synapse of additional receptor complexes, and with this greater transmitter binding to a given stimu­ lus, involves a process known as exocytosis (Carroll et al 2001). Additional receptors in the postsynaptic membrane is one of several possible changes in nervous system struc­ ture, and therefore function, that is clearly important for learning and certain types of memory. Another is the dis­ covery that following LTP postsynaptic structures form new synaptic connections with axons that contact them. A single axon can create multiple synapses on the same post­ synaptic neuron. It is likely that within minutes, or at the most hours, following tissue insult, activity initiated bio­ chemical events begin to bring about a change in nervous system anatomy as well as physiology. These mechanisms are likely to be initially responsible for clinically observed signs and symptoms but if maintained, they may also lay the foundation of long-term pain and disability (Carr & Goudas 1999). In their final expression, the processes of learning and memory involve long-lasting protein phosphorylation due to the action of persistently active kinases. Such long­ term physiological changes have the potential to become structurally permanent as a result of transcription medi­ ated protein synthesis. In this way, learning is facilitated and memory consolidated by the construction of addi­ tional entirely new synapses (Bear et aI 200l). In the pres­ ent context this would constitute a specific basis for long-term pain. Experiments using fear-learning para­ digms have demonstrated both AMPA receptor exocytosis and the synthesis of new protein contributing to the enhancement of LTP and long-term memory (Lin et al 2002, McKernan & Shinnick-Gallagher 1997, Scharf et al 2002). Experience with high 'emotional content' stimuli such as pain could consolidate 'laying down' of new synapses establishing a structural (as well as a physiolog­ ical) basis for chronic pain.

Differential sensitization

There has been a very strong emphasis on the role of the NMDA receptor in the central sensitization process. However, as noted above, it has now become apparent that activation of the NMDA receptor may not be critical to the development of all forms of central sensitization; the AMPA receptor also appears to be of critical importance. It has been suggested that the NMDA receptor is particularly important in relation to thermal sensitization and that it plays a lesser role in mechanical sensitization (Meller et al 1996). Co-activation of spinal AMPA and mGluR receptors induces an acute mechanical sensitization (Meller et al 1996). This is mediated through activation of phospholipase A2 leading to the production of arachidonic acid. It appears to be the products of the cyclooxygenase pathway for metabolism of arachidonic acid that are of most importance in generating mechanical sensitization (Meller et al 1996). Activation of NMDA receptors, phospholipase C, PKC and the production of nitric oxide appear to be less important factors in the development of mechanical sensitization and may be linked more to the development of thermal sensiti­ zation (Meller et aI 1996).
Trophic factors

As noted previously, BDNF is released centrally from a sub­ population of peripheral nociceptive neurons and has an important role in enhancing phosphorylation of NMDA receptors to maintain central sensitization (Boucher et al 2000). It appears to be particularly important in facilitating the development of thermal hyperalgesia, since intrathecal administration of the fusion molecule, tyrosine kinase receptor B - human immunoglobulin-y (trkB-IgG), signifi­ cantly reduces thermal hyperalgesia induced by carageenan inflammation (Boucher et a1 2000, Thompson et aI 1999). It also appears likely that GDNF may play a role in the devel­ opment of sensitization (Boucher et al 2000).
Neuroanatomical reorganization

Neuroanatomical reorganization of the laminar structure of the spinal cord is another important process that may con­ tribute to alterations in nociceptive system function. This mechanism appears to be particularly important when nerve injury has occurred. Under these circumstances, myelinated axons that normally terminate in laminae III and IV of the dorsal horn have been shown to sprout into lamina II of the dorsal horn, potentially developing synap­ tic connections with intrinsic neurons involved in the trans­ mission of nociceptive afferent inputs (Woolf et aI 1992). It has been postulated that this may constitute a mechanism whereby normally innocuous afferent input could con­ tribute to nociception (Woolf & Mannion 1999) and provide a neuroanatomical basis for the development of allodynia. Furthermore, it has been shown experimentally that in

1 62


some cases dorsal horn inhibitory interneurons are lost ('dark cells'). It appears that the destruction of these neu­ rons is a result of their susceptibility to a form of excitotox­ icity (Sugimoto et al 1990, Woolf & Salter 2000). Were this additional alterations in anatomy to also occur, the result is likely to be intense intractable pain and varying degrees of chronic functional disability. It should be emphasized that such changes in neuroanatomy are likely to be mainly con­ fined to situations in which the nervous system has been damaged and neuropathic pain has developed. Enhanced activation of central nervous system neurons is dependent on a range of mechanisms. It is apparent that cen­ tral sensitization and long-term potentiation in spinal cord neurons are relatively complex processes and that in com­ mon with peripheral processes, the nature of molecular changes underlying central sensitization may vary depend­ ing on the nature of the inducing stimulus. It is also apparent that other factors such as neuroanatomical reorganization may contribute to the changes that occur in central nervous system function post injury.

involvement of a number of key brain sites in pain percep­ tion. Some notable regions include: the anterior cingulate cortex (ACC), anterior insular cortex (IC), primary somatosensory cortex (51), secondary somatosensory cor­ tex (52), a number of regions in the thalamus and cerebel­ lum, and interestingly, areas such as the premotor cortex that are normally linked to motor function (Casey 1999). From this research there is abundant evidence of the dis­ tributed nature of the nociceptive system and the potential for close association between areas of the nervous system responding to pain and areas controlling motor function, and emotional state (Porro & Cavazzuti 1996). For example, it is clear that both the basal ganglia and the periaqueduc­ tal gray (PAG) region receive nociceptive inputs as well as coordinating important aspects of movement and motor control (Chudler & Dong 1995, Lovick 1991).
Attentional processes

Pain and nociceptive inputs can exert a strong influence on motor function and emotional state. As well as interactions at spinal cord level, integration of nociception and other central nervous system functions must occur at higher cen­ tres. It is also clear that pain perception can be strongly modulated by descending systems originating in various parts of the brain (Cervero & Laird 1996, Stamford 1995). This modulation can take the form of enhanced pain per­ ception, as well as the reduced pain perception associated with analgesic effects (Cervero & Laird 1996). It is now becoming apparent that as well as being influenced by pain, motor activity and emotional state can in turn influ­ ence pain perception (Dubner & Ren 1999). Consequently, the central nervous system is better viewed as an integrated cyclical system rather than the simple cause and effect sys­ tem enshrined in the distinction between afferent and effer­ ent aspects of function.
Functional brain imaging

Functional brain imaging studies are increasingly provid­ ing a means of bridging the gap between psychological studies and basic neurophysiological studies, and allowing us to gain some basic understanding of the way in which nociception is intimately integrated with many other aspects of central nervous system function. This work is providing insights into the complex ways in which cogni­ tive and emotional states can modulate pain perception. This is of considerable importance in understanding the influence of psychosocial factors on pain perception and pain report in the clinical situation. Studies investigating both experimentally induced pain and clinical pain states provide substantial evidence for the

Brain stem input from the ACC is particularly significant since the ACC has been invested with a pivotal role in integrating sensory and affective with attentional, cogni­ tive and emotional aspects of pain (Casey 1999, Davis et al 2000, Hutchison et al 1999, Kwan et al 2000, Price 2000, Rainville et aI 1997). This central role for the ACC has been demonstrated by using positron emission tomography (PET) as well as other imaging techniques (Fig. 12.3). PET is able to image changes in cerebral regional blood flow (rCBF) in response to a variety of noxious peripheral stim­ uli in awake human subjects (Casey 1999). Increases in rCBF responses during noxious stimulation are consid­ ered to reflect physiological changes in neuronal activity related to both nociceptive processing and the perception of pain. It is significant that, when methodological and analytical variations are taken into account, the same anatomical regions have been repeatedly highlighted across studies (Bushnell et al 1999, 2002, Casey 1999). Furthermore, the degree of rCBF response was found to correlate with reports of pain intensity in humans (Casey 1999, Hofbauer et al 2001). In addition to the ACC and IC, the most consistently acti­ vated supraspinal regions, using a variety of noxious stim­ uli, were motor centres, namely the premotor cortex and cerebellar vermis (Casey 1999). Evidence suggests that complimentary activation of cortical and subcortical motor areas is related to instructions for movements or postures intended to escape painful stimulation (Hsieh et al 1994, Price 2000). Converging on the ACC is information from higher centres (51, 52 via corticolimbic posterior parietal and insular cortices) that is considered to provide the organism with 'an overall sense of intrusion and threat to physical body and self' (Price 2000). This information is integrated with that from (pre)frontal cortical areas con­ cerned with future implications of the pain and with estab­ lishing response priorities. These include decisions and strategies to escape pain and any pain-evoking Situations.

N e u rophysiol ogy of pa i n a nd p a i n m od u l ation

1 63


. ______

Thermal Motor



D aACC D pACC pre·SMA D SMA proper

C -- Motor .------ Warm


. - - - - Heat pain VAC VPC

Cold pain

Figure 1 2.3 Spatial distribution of 95% confidence ellipses generated using the pooled activations of all subjects produced by: (A) the pain, thermal and motor tasks within the entire slice; (B) the pain, thermal and motor tas ks within the aSCC, pACC and areas of the SMA; and (C) warm, cool, cold pain, heat pain and motor tasks situated within the aACC, pACC and SMA. Reproduced from Kwan et al 2000. Key: (aACC anterior anterior cingulate cortex; pACC posterior anterior cingulate cortex; SMA supplementary motor area; VAC verti­ cal through anterior commissure; VPC vertical through posterior commissure.
= = = = =

Movement and postural planning concerned with the avoidance of pain obviously require intimate anatomical cooperation with pre and supplementary motor centres as well as close connections with centres of emotion and moti­ vation (Price 2000). When imaging pain patients, it is frequently found that the same sensory and motor cortical and limbic centres could be activated by normally non-painful, as well as noci­ ceptive stimuli (Bushnell et eI 2002). This helps confirm the presence of supraspinal neuron sensitization. Various types of chronic pain are also associated with extensive reorgani­ zation of sites encoding peripheral stimuli in the somatosensory cortex. In patients with back pain, for exam­ ple, in addition to a substantial increase in cortical activity to different intensities of peripheral cutaneous stimuli, there is a significant enlargement of the normal cortical rep­ resentation of the back. The expansion occurs medially into the neighbouring area normally representative of the leg and foot (Flor et al 1997). The additional neuronal activity

associated with an enlarged cortical representation may serve to enhance and maintain the pain experience (Flor et al 1997). Such changes are an important example of the degree of neuroplasticity that is likely to occur in the pres­ ence of chronic pain.

There is considerable overlap between the neuroanatomi­ cal and neurotransmitter systems modulating pain per­ ception and those controlling emotional state (Chapman 1996) . Bandler & Shipley (1994) describe a model of columnar organization that projects from regions of the frontal cortex, the hypothalamus, thalamus and amygdala to the PAG region of the mid-brain. These neuroanatomi­ cal connections may provide the basis for the interaction between cognitive and emotional states and pain percep­ tion, autonomic function and motor activity (Bandler & Shipley 1994).

1 64


Forebrain mediated modulatory systems

It is important to realise that central sensitization and long­ term potentiation in spinal cord neurons can also be brought about by activity of pain modulatory systems th�t descend to the dorsal horn of the spinal cord from the bram and, moreover, that these descending brain stem pain facil­ itatory systems are heavily connected with, and strongly influenced by, activity in key forebrain structures. Neuronal activity in forebrain structures related to cognitions and emotions could lead to an imbalance of descending modu­ latory systems. If that imbalance leads to an increase in endogenous facilitation, normally innocuous stimuli could be perceived as painful. It is possible that for some individ­ uals the diffuse nature and amplification of persistent pain may be in part the result of such an imbalance (see Fig. 12.4) (Dubner & Ren 1999). The influence attention and focusing can have on the perception of pain in humans was noted by Miron et al (1989). This research showed that, with contrived changes in directed attention, human volunteers reported alter­ ations in both the perceived intensity and unpleasantness, hence tolerance, of a transiently painful but non tissue dam­ aging thermal stimulus. These and other findings prompted Dubner & Ren (1999) to contend that the addi­ tion of a behavioural variable such as attention to a poten­ tially threatening stimulus results in sensitization of dorsal horn spinal cord neurons (see Fig. 12.5). In a series of behavioural studies incorporating electrophysiologically correlated data with a delayed-response task paradigm pri­ mates were rewarded for responding to a randomly deliv­ ered transient tissue threatening peripheral stimulus. It was found that expansion of receptive fields and increased responsiveness of second order trigeminal pain pathway neurons were directly related to the strength of engineered attention rather than stimulus intensity (Dubner & Ren 1999). Moreover, it was found that behavioural modulation

Figure 1 2.4 Per ip hera l nociceptive ('C') afferents sens it ize sp inal cord (SC) ne u rons as t hey ascend to s uprasp inal forebra in (FB) centres, on t he i r way to g iv in g excitato ry collatera ls to bra in stem in h ibito ry (iN) and facilitato ry ( FC) n u cle i. Pat hways from t hese rostral ventromed ial med ulla ry (RVM) n u cle i descend to mod ulate a ct iv ity in t he same SC ne urons. The IN and FC n ucle i are t hemselves contacted and may be infl uenced by a ct iv ity o ccu rring in t he FB ( cogn it ions, emot ions, atten­ tion). In this way fore bra in a ct iv ity can d im in is h or en hance sens it iza­ tion of SC ne urons w it h or w it ho ut ongo ing 'C' affe rent inp ut.

associated with selective attention to a perceived threat uti­ lizes the same forebrain and brain stem structures and mechanisms as are involved in the development, amplifica­ tion and maintenance of persistent pain following actual tissue damage and inflammation (Dubner & Ren 1999). The critical implication is that because there is a shared central pain producing and sustaining mechanism, the clinical con­ sequences of forebrain mediated selective attention a�e . likely to be functionally indistinguishable from those lill­ tially triggered by the primary peripheral pathology (Dubner & Ren 1999, Moog et aI 2002). It is apparent that there is considerable overlap between mechanisms responsible for both inhibition and facilitation of nociceptive inputs from spinal cord neurons. Depending on stimulus intensity both inhibition and facilitation of dor­ sal horn pain pathway activity could be achieved from many of the same brain stem nuclei, especially those located in the rostral ventromedial medulla (RVM) (Urban & Gebhart 1999). Interestingly, despite this apparent anatomical overlap, the opposing neurophysiological con­ sequences (inhibition, facilitation) have been shown to involve different spinal cord pathways and neurotransmit­ ters, and to be dorsal horn lamina- and receptor-specific (Urban & Gebhart 1999). It has also been shown that simul­ taneously lesioning nominally inhibitory (nucleus raphe magnus, NRM) and facilitatory (nucleus reticularis gigan­ tocellularis, NGe) sites completely reversed their custom­ ary (opposing) effects on spinal cord neurons (Wei et al 1999a). The responses of wide dynamic range dorsal horn neu­ rons (laminae I-VI) to mechanical peripheral stimuli can also be modulated by electrical and chemical (glutamate) stimulation applied at a range of sites in the rostral medial medulla (RMM) (Zhuo & Gebhart 2002). The often biphasic changes were specific for the site of RMM stimulation and not the particular neuron from which recordings were taken. Thus, individual dorsal horn neuron responses to noxious peripheral stimuli are enhanced by stimulation at some, and inhibited by stimulation at other, RMM sites. At some RMM sites, activity in the same neuron was facilitated with lower and inhibited with higher intensity stimulation in a similar manner to RVM nuclei. Significantly, similar findings were obtained for responses evoked in dorsal horn neurons by both noxious and non-noxious mechanical peripheral stimulation. Together with other results, such findings led the authors to conclude that spinal transmission of noxious and non­ noxious peripheral mechanical stimuli 'is subject to descending influences, including facilitatory influences that may contribute to exaggerated responses ... in some chronic pain states' (Zhuo & Gebhart 2002). In this regard, certain brain stem nuclei have been iden­ , tified as sources of potential tonic facilitation and persistent pain. These include the nucleus gigantocellularis, nucleus gigantocellularis pars alpha (Dubner & Ren 1999� and �e dorsal reticular nucleus of the medulla (Lima & AlmeIda

Neurophysio logy of p a i n a nd p a i n modu l a tion

1 65

2002). Descending facilitation is implicated in the develop­ ment of centrally created secondary hyperalgesia associ­ ated with inflammatory and neuropathic pain (Porreca et al 2002). It is suggested that the reason for the existence of such a system is that descending facilitation initially has a discrete protective function, especially with inflammatory pain conditions (Porreca et al 2002). By shifting the balance in favour of facilitation, forebrain activity related to attention and evaluation of threat could have a role in the initiation and maintenance of central sen­ sitization in spinal cord neurons (Fig. 12.4). It might be anticipated that this would lead to clinically observed symptoms and signs that would be difficult to distinguish from symptoms and signs occurring as a result of periph­ eral injury. Understanding this is of considerable impor­ tance in clinical reasoning and the identification of patients where forebrain mediated central sensitization is strongly suspected of being the cause, or major component, of their prolonged pain and functional disability. Certain cognitive styles have been associated with gross amplification of pain and its extension in the absence, or beyond the period for healing, of tissue damage (Aronoff 1998, Bacon et al 1994, Bass 2000, Crombez et al 1999, Ferrari & Schrader 2001, Jensen et al 1994, Keogh et a1 2001, Linton 2000, Sullivan et al 2001, Vlaeyen & Linton 2000, Waddell et al 1993). These include somatization, catastrophizing, and hypervigilance. Figure 12.5 provides a model of the pos­ sible relationship between these factors and the develop­ ment of chronic pain. Forebrain mediated central sensitization may provide a neuronal mechanism whereby such cognitive styles can contribute to up-regulation of the nociceptive system.
Loss of inhibition

predisposition: Catastrophizing JExaggerated threat value of pain (Appraisal model)

JFear Need for early detection JSelective attention - hypervigilance JPain intensification reinforces beliefs, fears, attributions JAvoidance leads to disuse, depression, disability JSomatic preoccupation rumination, magnification JM isinterpretation of irrelevant sensations increased protection, 'focussing' J'Activity intolerance'
negative self-efficacy beliefs

external locus of control-'passive coping style'

You fix m e !
Figu re 1 2 . 5 Model of the relationship between so me ident if ied maladapt ive cogn it ive processes and the develop ment of ch ron ic pa in. Note the p ivotal role of select ive attent ion - hyperv ig ilance.

On balance, the net effect of descending brain stem systems on spinal cord neurons, normally as well as under routine inflammatory conditions, seems to be inhibitory. However, it is possible for various factors to reduce this inhibition and shift the balance in favour of facilitation. Experimental evidence for the importance of blockade of descending inhibitory pathways in inducing central sensiti­ zation include the observation that bilateral lesions of the dorsolateral funiculus in the rat led to a significant decrease in the latency for paw withdrawal to a noxious stimulus (Wei et al 1999b). Dorsolateral funiculi appear to be a pre­ ferred pathway for descending pain inhibitory systems (Dubner & Ren 1999, Urban & Gebhart 1999). Similarly, temporary spinal cord block (lidocaine (lignocaine)) caused dorsal hom nociceptive specific and wide dynamic range neurons to expand their receptive fields and increase their responsiveness to afferent input. These effects were further enhanced in experimentally 'inflamed' animals (hindpaw injection of complete Freund's Adjuvant or carrageenan) (Ren & Dubner 1996). Selective anaesthesia (lidocaine (lig­ nocaine)) of the nucleus raphe magnus (NRM), in the RVM

caused nociceptive-specific neurons in the spinal cord to increase their background discharge, expand their receptive fields and respond in an exaggerated and abnormal manner to subsequent peripheral stimuli (Ren & Dubner 1996). In addition, selective chemical lesion of NRM 5HT-containing neurons in experimentally 'inflamed' animals resulted in demonstrable behavioural hypersensitivity. This was accompanied by the appearance of the Fos protein bilater­ ally in all laminae of the animals' spinal cord (Wei et al 1998). The Fos protein is a gene transcription by-product and recognized biological marker for enhanced neural activity (Menetrey et al 1989). Similar effects were observed in superficial laminae following lesions of an inhibitory noradrenergic locus coeruleus-dorsal hom pathway. The foregoing provides evidence for the importance of brain stem descending inhibitory systems in regulating the excitability of spinal cord pain pathway neurons (Wei et al 1998). Moreover, it suggests that disruption of one or more of the elements of this system can result in, among other things, the equivalent of central sensitization.

1 66



Among brain stem nuclei so far identified as the origin of descending (or locally acting) pain pathway facilitatory systems is the nucleus reticularis gigantocellularis (NGC). Low intensity stimulation in the RVM at or near the NGC has been shown to cause lingering excitation of some spinal cord spinothalamic tract neurons as well as a decrease in the latency of the tail-flick response (Haber et al 1980, Zhuo & Gebhart 1990). Stimulation of the NGC also enhanced the responses of primate spinothalamic tract neurons to tran­ sient noxious stimuli (Haber et al 1980). Selective lesions (ibotenic acid) of NGC in the 'inflamed' rat led to a signifi­ cant increase in the latency for the paw withdrawal reflex and a marked reduction in the presence of Fos protein bilat­ erally in all laminae of the spinal cord. Such experiments provide evidence that NGC is capable of enhancing and/ or maintaining central sensitization at the spinal cord level (Dubner & Ren 1999). Significantly, secondary hyperalgesia was completely blocked by ibotenic acid lesions of the medulla that included NGC (Urban et al 1996, 1999). Almeida et al (1999) confirmed the presence of a noci­ ceptively driven pathway from the dorsal reticular nucleus (DRt) of the medulla to superficial and deeper laminae of the dorsal horn of the spinal cord. Importantly, the findings were backed up by calculating the number of cells express­ ing noxiously induced c-fos in both superficial and deeper laminae of the spinal cord. The Fos protein is known to be a reliable marker of neuronal (hyper)activity (Menetrey et aI 1989). Because it has every appearance of being rever­ beratory or self-sustaining, the DRt-dorsal horn circuit could be particularly significant, not only with respect to pain amplification but also in terms of chronicity. Fields et al characterized specific groups of cells in RVM that may form the basis for bidirectional control (Fields et al 1983). They identified three distinct cell groups that show characteristically different responses during a thermal tail­ flick test in the rat. The cells were described as off cells which show continuous ongoing activity and pause just before the tail-flick response occurs, on cells that are toni­ cally inhibited but display a pronounced burst of activity just before the tail-flick response occurs and neutral cells whose activity is not specifically related to the pain response (Fields et aI 1983). This group hypothesized that off cells provide a tonic inhibition of nociceptive transmission cells in the spinal cord that is disinhibited and then facilitated by on cell activ­ ity when pain occurs. Both types of neuron descend to appropriate laminae (I, II, IV) of the dorsal horn of the spinal cord and both may be influenced by stimulation of the PAG. Interestingly, Wei et al (1999a) recently demonstrated that the function of inhibitory 'off' cells may be suppressed (via inhibitory interneurons) by activity in nearby NGC neurons. Thus, NGC neurons probably exert their dorsal horn effects through several circuits. However, one mechanism for

endogenous descending facilitation, or sensitization, of spinal cord pain pathway neurons could be the local inhi­ bition of a brain stem descending inhibitory system. The net effect of stimulation of NGC is facilitation of central sen­ sitization at the spinal cord level (Ren et aI 2000). It is pos­ sible that this facilitation may be initiated or controlled by the anterior cingulate cortex since electrical and chemical stimulation of ACC elicits marked facilitation of the tail­ flick reflex to noxious thermal stimulation (Calejesan et al 2000). This effect is apparently mediated through relays in PAG and RVM . There is now substantial evidence for pathways from acknowledged rostral 'pain relevant' cortical and subcorti­ cal centres to both the PAG and RVM. There is also com­ pelling behavioural evidence that these forebrain centres are capable of exerting powerful clinically significant influ­ ences on various nuclei located within brain stem struc­ tures, including NGC . Together this anatomical, physiological and behavioural evidence helps confirm the long recognized, critical influence that such forebrain prod­ ucts as cognitions, emotions, attention and motivation have on the clinical pain experience. The evidence further endorses detrimental effects that recognized psychosocial factors have on motor control and adaptive function, both spontaneously and as a result of pain.

The major consequences of molecular changes in spinal cord neurons are increased synaptic efficacy and increased neuronal excitability. Neuronal plasticity leading to increased synaptic efficacy and increased neuronal excitability in spinal cord neurons conveying nociceptive information is also likely to influence activity in other neu­ ronal pools with which the central nociceptive neurons make synaptic connections. This could account for changes in motor system function, which are clinical features of many pain states.
Enhanced withdrawal reflexes

It is clear that this hyperactive state of spinal cord neurons is associated with important changes in terms of sensori­ motor function. In his seminal study, Woolf showed that the establishment of central sensitization is associated with facilitation of flexor withdrawal reflex responses (Woolf 1984). A prolonged increase in the response duration is maintained for several days and in some cases may still be present weeks later when tissue healing is presumed to have occurred (Woolf 1984). Altered flexor withdrawal reflexes may be of clinical importance in tests such as the straight leg raise �d the brachial plexus tension test. Increased muscle activIty has been demonstrated in normal subjects when undergoing neural tissue provocation tests (Balster & Jull 1997). These studies support the proposal that muscle activitY protects

N e u rop hysiology of p a i n a n d p a i n mod u l ation

1 67

the nervous system from tensile forces. It has been sug­ gested that this increase in muscle activity is due to activa­ tion oHhe flexor withdrawal reflex (Hall et a1 1998, Wright et aI 1994). Hall et al (1998) showed that flexor muscle activ­ ity is more easily elicited in chronic pain patients than in normal volunteers during the straight leg raise test (Hall et aI 1998).
Vicious cycle model

In addition to increased muscle activation attributable to the influence of pain and tissue damage on alpha motoneu­ ron function, it has been suggested that pain may influence the excitability of gamma motoneurons contributing to the development of increased muscle tension or spasm. The vicious cycle model is often alluded to in the literature. As outlined by Johansson & Sojka (1991), the basic concept is that stimulation of nociceptive afferents from muscles excites dynamic and static fusimotor neurons enhancing the sensitivity of primary and secondary muscle spindle affer­ ents. Increased activity of primary muscle spindle afferents increases muscle stiffness. This increased muscle stiffness then leads to increased metabolite production and, follow­ ing the vicious cycle formula, a further increase in muscle stiffness. In addition, increased activity in the secondary spindle afferents projects back onto the gamma system per­ petuating enhanced muscle stiffness. These effects are thought to be important in generating muscle spasm and pain Gohansson & Sojka 1991). There are several studies demonstrating enhanced activ­ ity in primary and secondary spindle afferents following the application of chemical mediators such as potassium chloride, lactic acid, bradykinin and serotonin (Djupsjobacka et a1 1995, Johansson et aI 1993). In addition to altered responses following local muscle injection, these researchers have also demonstrated modulation of second­ ary muscle spindle afferents following injection of bradykinin into the contralateral muscle (Djupsjobacka et al 1995). This model may provide some explanation of muscle spasm when it is a significant component of the clinical presentation. However, it provides very little explanation of situations in which we see muscle inhibition and wastage as a result of pain and a number of studies have failed to show an increase in resting EMG activity as might be pos­ tulated by this model.
Pain adaptation model

facilitation of antagonist motor units. This leads to an over­ all limitation of movement in any desired direction. The proposed alterations in neural function would be manifest as a reduction in the ability to activate the agonist muscle, a time delay in activating the agonist muscle and a reduction in the maximum force output from the agonist muscle. Increased activity in antagonist muscles and a delay in pro­ ducing reciprocal inhibition of these muscles might also be anticipated. Movement becomes slower, muscles appear to be weaker and the overall range of movement accom­ plished may be reduced (Lund et al 1991 ). Deficits of this type have been demonstrated in patients with low back pain and in normal subjects following the injection of hypertonic saline into the lumbar paraspinal muscles (Arendt-Nielsen et a1 1996) and the muscles of mastication (Svensson et al 1996). This model may represent a good explanation of the limitation of movement that occurs in the acute pain situation. It is apparent, however, that motor dysfunction in chronic pain states may be a somewhat more complex phenomenon.
Emerging models

Lund and colleagues refuted the vicious cycle model and suggested that pain reduces the ability to contract muscles rather than making them hyperactive (Lund et al 1991). Their model, termed the pain adaptation theory, is strongly linked to the phenomenon of central sensitization. They propose that the effect of noxious stimulation is to alter the activity of type II spinal cord interneurons such that there is increased inhibition of agonist motor units and increased

Over the last decade researchers have begun to investigate the influence of pain on patterns of neuromuscular activa­ tion and control. It has been suggested that the presence of pain leads to inhibition or delayed activation of muscles or muscle groups that perform key synergistic functions to limit unwanted motion (Sterling et al 2002). This produces alterations in the patterns of motor activity and recruitment during functional movement. It has been suggested that this inhibition usually occurs in the deep muscles, local to the involved joint, that perform a synergistic function in order to control joint stability (Hides et al 1996, Hodges & Richardson 1996, Voight & Wieder 1991). In both the lumbar and the cervical spine, the dysfunc­ tional muscles appear to be the deep muscles that attach directly to the vertebrae. These muscles span the vertebrae and perform important synergistic functions to stabilize articular segments, rather than being primarily responsible for movement production (Cholewicki et al 1997). It appears that while changes in the control of these muscles may be initiated in the presence of pain and tissue injury, they are often sustained beyond the acute pain phase and may contribute to the chronicity of many musculoskeletal problems. It is clear that pain can produce many changes in motor activity. Some of these changes can be explained by periph­ eral mechanisms in the muscles themselves and by mecha­ nisms within the central nervous system. Certainly, pain has a potent effect on motor activity and control. The dysfunc­ tion that occurs in the neuromuscular system in the presence of pain is complex. In addition to the more obvious changes, such as increased muscle activity in some muscle groups and inhibition of others, more subtle anomalous patterns of neuromuscular activation appear to occur. Some elements of

1 68


both the vicious cycle and pain adaptation models may be important in acute and chronic pain states. However, nei­ ther of these models can fully explain the prolonged changes in motor function that are seen after tissue injury. Loss of selective activation and inhibition of certain muscles that perform key synergistic functions, leading to altered patterns of neuromuscular activation, and the ensuing loss of joint stability and control are initiated with acute pain and tissue injury. However, these phenomena persist and could be one reason for chronic symptoms.

Recent advances in our knowledge of pain have provided a much greater insight into the many ways in which nocicep­ tive system activity can be enhanced in response to tissue injury. It is clear that both peripheral and central mecha­ nisms are important and that subtle variations in the mech­ anisms activated can result in different forms of altered sensitivity being induced. Mechanisms exist to sensitize nociceptors, to recruit previously inactive nociceptors and to utilize afferent inputs via myelinated neurons to con­ tribute to nociception. These mechanisms contribute to sub­ stantial spatial and temporal summation of nociceptive inputs. Central mechanisms appear to be particularly important in controlling the spread of sensitivity to unin­ jured tissues surrounding the region of tissue damage.

Rapidly developing areas of research are also improving our knowledge of the interrelationship between pain, motor function and emotional state. We are beginning to move away from a largely peripheralist view of tissue injury to a much more integrated understanding of the influence of pain and injury on the central nervous system and the patient as a whole. This encompasses an emerging view of the nociceptive system as a highly distributed sys­ tem that interacts with many other neuronal systems. The immense impact of pain and tissue injury on the central nervous system and the plasticity induced by the presence of pain are becoming increasingly apparent. It is now clear that the relative setting of descending facilitatory and inhibitory projections from brain to spinal cord may signif­ icantly affect the perceptions of nociceptive inputs. Cognitions, emotions and attentional state can all have important influences on the balance between descending inhibition and facilitation. Ultimately, improved under­ standing of these mechanisms should lead to the develop­ ment of a more comprehensive approach to the management of patients with pain.
KEYWOR DS musculoskeletal pain periphera l sensitization centra I sensitization pain inhibitory mecha nisms

Almeida A, Storkson R, Lima D, Hole K, Tjolsen A 1999 The medullary dorsal reticular nucleus facilitates pain behaviour induced by formalin in the rat. European Journal of Neuroscience 1 1 : 110-122 Anand P 1995 Nerve growth factor regulates nociception in human health and disease. British Journal of Anaesthesia 75: 201-208 Anbar M, Gratt B M 1997 Role of nitric oxide in the phYSiopathology of pain. Journal of Pain Symptom Management 14: 225-254 Arendt-Nielsen L, Graven-Nielsen T, Svarrer H, Svensson P 1996 The influence of low back pain on muscle activity and coordination during gait: a clinical and experimental study. Pain 64: 231-240 Aronoff G M 1998 Myofascial pain syndrome and fibromyalgia: a critical assessment and alternate view. Clinical Journal of Pain, 14: 74-85 Bacon N M, Bacon S F, Atkinson J H, et al 1994 Somatization symptoms in chronic low back pain patients. Psychosomatic Medicine 56: 1 1 8-127 Balster S M, Jull G A 1997 Upper trapezius muscle activity during the brachial plexus tension test in asymptomatic subjects. Manual Therapy 2: 144-149 Bandler R, Shipley M T 1994 Columnar organization in the midbrain periaqueductal gray: modules for emotional expression? Trends in Neurosciences 17: 379-389 Bass C 2000 Somatization. Medicine 28(5): 68-71 Bear M F, Connors B W, Paradiso M A 2001 Neuroscience: exploring the brain. Lippincott, Philadelphia, chs 23-24 Bevan S 1996 Signal transduction in nociceptive afferent neurons in inflammatory conditions. In: Kumazawa T, Kruger L, Mizumura K (eds) Progress in brain research. Elsevier Science BV, Amsterdam, vol 113, pp 201-213 Boucher T

1, Kerr B I,

Ramer M S, Thompson S W N, McMahon S B

2000 Neurotrophic factor effects on pain-signaling systems. In: Devor M, Rowbotham M C, Wiesenfeld-Hallin Z (eds) Proceedings of the 9th World Congress on Pain, Progress in Pain Research and Bushnell M C, Duncan G H, Hofbauer R K, Ha B, Chen J I, Carrier B 1999 Pain perception: is there a role for primary somatosensory cortex? Proceedings of the National Academy of Sciences of the United States of America 96: 7705-7709 Bushnell M K, Villemure C, Strigo I, Duncan G H 2002 Imaging pain in the brain: the role of the cerebral cortex in pain perception Calejesan A A, Kim S I, Zhuo M 2000: Descending facilitatory and modulation. Journal of Musculoskeletal Pain 10( 1 / 2): 59-72 modulation of a behavioural nociceptive response by stimulation in the adult rat anterior cingulate cortex. European Journal of Pain 4: 83-96 Carr D B, Goudas L C 1999 Acute pain. Lancet 353: 2051-2058 Carroll R C, Beattie E C, von Zastrow M, Malenka R C 2001 Role of AMPA receptor endocytosis in synaptic plasticity. Nature Reviews Neuroscience 2: 315-324 Casey K L, 1999 Forebrain mechanisms of nociception and pain: analysis through imaging. Proceedings of the National Academy of Sciences of the United States of America 96: 7668-7674 Cervero F, Laird Management. IASP Press, Seattle, vol 16, pp 1 75-189

I M A 1996 From acute to chronic pain: mechanisms

and hypotheses. In: Carli G, Zimmerman M (eds) Progress in brain research. Elsevier Science BV, Amsterdam, vol 110, pp 3-15 Cesare P, Moriondo A, Vellani V, McNaughton P A 1999 Ion channels the United States of America 96: 7658-7663 gated by heat. Proceedings of the National Academy of Sciences of -

N e u rop hysiology of p a i n a n d p a i n mod u l ation

1 69

Chapman C R 1996 Limbic processes and the affective dimension of pain. In: Carli G, Zimmerman M (eds) Progress in brain research. Else'lier Science BV, Amsterdam, vol 110, pp 63-81 Chole�icki J, Panjabi M M, Khachatryan A 1997 Stabilizing function of

Harding V 1999 The role of movement in acute pain. In: Max M (ed) Pain 1999 - an updated review. IASP Press, Seattle, pp 159-169 Hardy J 0, Wolff H G, Goodell H 1950 Experimental evidence of the nature of cutaneous hyperalgesia. Journal of Clinical Investigation 29: 115-140 Hides J A, Richardson C A, Jull G A 1996 Multifidus muscle recovery is not automatic following resolution of acute first episode low back pain. Spine 21: 2763-2769 Hodges P W, Richardson C A 1996 Insufficient muscular stabilisation of the lumbar spine associated with low back pain: a motor control examination of transversus abdominus. Spine 21: 2640-2650 Hofbauer R K, Rainville P, Duncan G H, Bushnell M C 2001 Cortical representation of the sensory dimension of pain. Journal of Neurophysiology 86: 402-411 Hsieh J C, Hagermark 0, Stahle-Backdahl M et al 1994 Urge to scratch represented in the human cerebral cortex during itch. Journal of Neurophysiology 72: 3004-3008 Hutchison W 0, Davis K 0, Lozano A M, Tasker R R, Dostrovsky J ° 1999 Pain-related neurons in the human cingulate cortex. Nature Neuroscience 2: 403-405 Jensen M P, Turner J A, Romano J M, Lawler B K 1994 Relationship of pain-specific beliefs to chronic pain adjustment. Pain 57: 301-309 Johansson H, Sojka P 1991 Pathophysiological mechanisms involved in genesis and spread of muscular tension in occupational muscle pain and in chronic musculoskeletal pain syndromes: a hypothesis. Medical Hypotheses 35: 196-203 Johansson H, Djupsjobacka M, Sjolander P 1993 Influences of the gamma-muscle spindle system from muscle afferents stimulated by KCL and lactic acid. Neuroscience Research 16: 49-57 Keogh E, Ellery D, Hunt C, Hannent I 2001 Selective attentional bias for pain-related stimuli amongst pain fearful individuals. Pain 91(1 /2): 91-100 Kingsley R E 2000 Concise text of neuroscience. Lippincott, Philadelphia, pp 102-103 Kumazawa T 1996 The polymodal nociceptor: bio-warning and defense system. Progress in Brain Research 113: 3-18 Kwan C L, Crawley A P, Mikulis D J, Davis K D 2000 An fMRI study of the anterior cingulate cortex and surrounding medial wall activations evoked by noxious cutaneous heat and cold stimuli. Pain 85: 359-374 Levine J D, Fields H L, Basbaum A I 1993 Pep tides and the primary afferent nociceptor. Journal of Neuroscience 13: 2273-2286 Lima D, Almeida A 2002 The medullary dorsal reticular nucleus as a pronociceptive centre of the pain control system. Progress in Neurobiology 66: 81-108 Lin B, Brucher F A, Colgin L L, Lynch 2002 Long-term potentiation alters the modulator pharmacology of AMPA-type glutamate receptors. Journal of NeurophYSiology 87: 2790-2800 Linton S J 2000 A review of psychological risk factors in back and neck pain. Spine 25: 1148-1156 Lovick T A 1991 Interactions between descending pathways from the dorsal and ventrolateral periaqueductal gray matter in the rat. In: Depaulis A, Bandlier R (eds) The midbrain periaqueductal gray matter. Plenum Press, New York, pp 101-120 Lund J P, Donga R, Widmar C G, Stohler C S 1991 The pain adaptation model: a discussion of the relationship between chronic musculoskeletal pain and motor activity. Canadian Journal of Physiology and Pharmacology 69: 683-694 Ma Q P, Woolf C J 1995 Noxious stimuli induce an N-methyl-D­ aspartate receptor-dependent hypersensitivity of the flexion withdrawal reflex to touch: implications for the treatment of mechanical allodynia. Pain 61: 383-390 Ma Q-P, Woolf C J 1996 Progressive tactile hypersensitivity: an inflammation-induced incremental increase in the excitability of the spinal cord. Pain 67: 97-106

trunk flexor-extensor muscles around a neutral spine posture. Spine
22: 2207-2212. Chudler E H, Dong W K 1995 The role of the basal ganglia in nociception and pain. Pain 64: 3-38 Coderre T J, Melzack R 1987 Cutaneous hyperalgesia: contributions of the peripheral and central nervous systems to the increase in pain sensitivity after injury. Brain Research 404: 95-106 Cook A J, Woolf C J, Wall P 0, McMahon S B 1987 Dynamic receptive field plasticity in rat spinal cord dorsal horn following C primary afferent input. Nature 325: 151-153 Crombez G, Eccleston C, Baeyens F et al 1999 Attention to chronic pain is dependent upon pain-related fear. Journal of Psychosomatic Research 47(5): 403-410 Davis K 0, Taub E, Duffner F et al 2000 Activation of the anterior cingulate cortex by thalamic stimulation in patients with chronic pain: a positron emission tomography study. Journal of Neurosurgery 92: 64-{j9 Dickenson A H 1995 Central acute pain mechanisms. Annals of Medicine 27: 223-227 Djupsjobacka M, Johansson H, Bergenheim M, Wenngren B I 1995 Influences on the gamma-muscle spindle system from muscle afferents �timulated by increased intramuscular concentrations of bradykinin and 5-HT. Neuroscience Research 22: 325-353 Dray A 1995 Inflammatory mediators of pain. British Journal of Anaesthesia 75: 125-131 Dray A 1996 Neurogenic mechanisms and neuropeptides in chronic pain. In: Carli G, Zimmerman M (eds) Progress in brain research. Elsevier, Amsterdam, vol 110, pp 85-94 Dubner R, Ren K 1999 Endogenous mechanisms of sensory modulation. 6(Suppl. ): 545-53 Duggan A W, Hendry I A, Morton C R, Hutchinson W 0, Zhao Z Q 1988 Cutaneous stimuli releasing immunoreactive substance P in the dorsal horn of the cat. Brain Research 451: 261-273 Duggan A W, Hope P J, Jarrot B, Schaible H-G, Fleetwood-Walker S M 1990 Release, spread, and persistence of immunoreactive neurokinin A in the dorsal horn of the cat following noxious cutaneous stimulation: studies with antibody microprobes. Neuroscience 35: 195-202 Ferrari R, Schrader H 2001 The late whiplash syndrome: a biopsychosocial approach. Journal of Neurology, Neurosurgery and Psychiatry 70: 722-726 Fields H L, Bry J, Hentall 1, Zorman G 1983 The activity of neurons in the rostral medulla of the rat during withdrawal from noxious heat. Journal of Neuroscience 3: 2545-2552 Flor H, Braun C, Elbert T, Birbaumer N 1997 Extensive reorganisation of primary somatosensory cortex in chronic back pain patients. Neuroscience Letters 224: 5-8 Gordh T, Karlsten R, Kristensen J 1995 Intervention with spinal NMDA, adenosine, and NO systems for pain modulation. Annals of Medicine, 27: 229-234 Haber L H, Martin R F, Chung J M, Willis W 0 1980 Inhibition and excitation of primate spinothalamic tract neurons by stimulation in region of nucleus reticularis gigantocellularis. Journal of Neurophysiology 43: 1578-1593 Hall T, Zusman M, Elvey R 1998 Adverse mechanical tension in the nervous system? Analysis of straight leg raise. Manual Therapy 3: 140-146 Handwerker H 0, Reeh P W 1991 Pain and inflammation. In: Bond M R, Charlton I E, Woolf C J (eds) Pain research and clinical management. Proceedings of the VIth World Congress on Pain. Elsevier, Amsterdam, pp 59-70 Handwerker H 0, Reeh P W 1992 Nociceptors, chemosensitivity and sensitization by chemical agents. In: Willis W 0 (ed) Hyperalgesia and allodynia. Raven Press, New York, pp 107-115

1 70


McKernan M G, Shi.n.nick-Galagher P 1997 Fear conditioning induces a 607-611 lasting potentiation of synaptic currents in vitro. Nature 390:

Scharf M T, Woo N H, Lattal K M et al 2002 Protein synthesis is

required for the enhancement of long-term potentiation and long­ 2770-2777

McMahon S B, Bennett D L, Priestley J V, Shelton D L 1995 The

term memory by spaced learning. Journal of Neurophysiology 87:

biological effects of endogenous nerve growth factor on adult Medicine 1 : 774-780

sensory neurons revealed by a trkA-IgG fusion molecule. Nature

Schmidt R F 1996 The articular polymodal nociceptor in health and

disease. In: Kumazawa T, Kruger L, Mizumura K. (eds) Progress in brain research. Elsevier Science BV, Amsterdam, vol 113, pp 53-81

Mao J, Price D D, Mayer D J 1995 Mechanisms of hyperalgesia and Pain 62: 259-274

morphine tolerance: a current view of their possible interactions.

Shu X Q, Mendell L M 1999 Neurotrophins and hyperalgesia. States of America 96: 7693-7696

Proceedings of the National Academy of Sciences of the United

Mayer E A, Gebhart G F 1994 Basic and clinical aspects of visceral Meller S T, Gebhart G F 1993 Nitric oxide (NO) and nociceptive Meller S T, Dykstra C, Gebhart G F 1996 Acute mechanical hyperalgesia AMPA and metabotrophic glutamate receptors, activation of in the rat can be produced by coactivation of spinal ionotrophic processing in the spinal cord. Pain 52: 127-136 hyperalgesia. Gastroenterology 107: 271-293

Siddall P J, Cousins M J 1998 Introduction to pain mechanisms:

implications for neural blockade. In: Cousins M J, Bridenbaugh P O pain. Lippincott-Raven, Philadelphia Anaesthesia 75: 21 7-227 (eds) Neural blockade in clinical anesthesia and management of

Stamford J A 1995 Descending control of pain. British Journal of Steen K H, Reeh P W, Anton F, Handwerker H 0 1992 Protons

Menetrey D, Gannon A, Levine J D, Basbaum A I 1989 Expression of c­ cord in response to noxious somatic, articular, and visceral fos protein in interneurons and projection neurons of the rat spinal stimulation. Journal of Comparative Neurology 285: 1 77-195

Science BV, Amsterdam, vol 110, pp 1 77-192

Carli G, Zimmerman M (eds) Progress in brain research. Elsevier

phospholipase A2 and generation of cyclooxygenase products. In:

selectively induce lasting excitation and sensitisation to mechanical stimuli of nociceptors in rat skin, in vivo. Journal of Neuroscience 12: 86--95.

Sterling M, Jull G, Wright A 2001 The effect of musculoskeletal pain on Sugimoto T, Bennett G J, Kajander K C 1990 Transsynaptic motor activity and control. Journal of Pain 2: 135-145

Merskey H, Bogduk N 1994 Classification of chronic pain: descriptions Press, Sea ttle of chronic pain syndromes and definitions of pain terms. !ASP

degeneration in the superfiCial dorsal hom after sciatic nerve injury: effects of a chronic constriction injury, transection, and strychnine. Pain 42: 205-213

Miron D, Duncan G H, Bushnell M K 1989 Effects of attention on 345-352 the intensity and unpleasantness of thermal pain. Pain 39:

Sullivan M J, Thorn B, Haythornthwaite J A et al 2001 Theoretical Clinical Journal of Pain 17: 52-64

perspectives on the relation between catastrophizing and pain.

Mizamura K, Kumazawa T 1996 Modification of nociceptor response by inflammatory mediators and second messengers implicated in their action: a study in canine testicular polymodal receptors. In: research. Elsevier Science BV, Amsterdam, vol 1 1 3, pp 115-141 Kumazawa T, Kruger L, Mizumura K (eds) Progress in brain

Svensson P, Arendt-Nielson L, Houe L 1996 Sensory-motor interactions of human experimental jaw muscle pain: a quantitative analysis. Pain 64: 241-250

Thacker M, Gifford L 2002 A review of the physiotherapy management in pain 3. CNS Press, Falmouth, pp 119-141

of complex regional pain syndrome. In: Gifford L (ed) Topical issues

Moog M, Quintner J, Hall T, Zusman M 2002 The late whiplash 283-294

syndrome: a psychophysical study. European Journal of Pain 6:

Thompson S W, Bennett D L, Kerr B J, Bradbury E J, McMahon S B 1999 nociceptive responses in the spinal cord. Proceedings of the National Torebjork E, Lundberg L, La Motte R 1992 Central changes in the processing of mechanoreceptive input in capsaicin-induced 448: 765-780 Academy of Sciences of the United States of America 96: 7714-7718 Brain-derived neurotrophic factor is an endogenous modulator of

Neumann S, Doubell T P, Leslie T A et al 1996 Inflammatory pain primary sensory neurones. Nature 384: 360-364

hypersensitivity mediated by phenotypic switch in myelinated

Porreca F, Ossipov M H, Gebhart G F 2002 Chronic pain and medullary Porro C A, Cavazzuti M 1996 Functional imaging of the pain system in research. Elsevier Science BV, Amsterdam, vol 110, pp 47--{)2 dimension of pain. Science 288: 1 769-1772 descending facilitation. Trends in Neuroscience 25: 319-325

secondary hyperalgesia in humans. Journal of Physiology (London) Urban M 0, Gebhart G F 1999 Supraspinal contributions to the United States of America 96: 7687-7692

man and animals. In: Carli G, Zimmerman M (eds) Progress in brain

hyperalgeSia. Proceedings of the National Academy of Sciences of

Price D D 2000 PsycholOgical and neural mechanisms of the affective Rainville P, Duncan G H, Price D D, Carrier B, Bushnell M C 1997 Pain affect encoded in human anterior cingulate but not somatosensory cortex. Science 277: 968-971 Anaesthesia 75: 145-156

Urban M 0, Jiang M C, Gebhart G F 1996 Participation of central descending nociceptive facilitatory systems in secondary hyperalgesia produced by mustard oil. Brain Research 737: 83-91

Vlaeyen J W, Linton S J 2000 Fear-avoidance and its consequences in Voight M L, Wieder D L 1991 Comparative reflex response times of subjects with extensor mechanism dysfunction: an 131-137 chronic musculoskeletal pain: a state of the art. Pain 85: 317-332

Rang H P, Urban L 1995 New molecules in analgesia. British Journal of Reeh P W, Steen K H 1996 Tissue acidosis in nociception and pain. In: Kumazawa T, Kruger L, Mizumura K (eds) Progress in brain research. Elsevier Science BV, Amsterdam, vol 113, pp 143-151 in rats with persistent hind paw inflammation. Journal of Neurophysiology 76: 3025-3037

vastus medialis obliquus and vastus lateralis in normal subjects and electromyographic study. American Journal of Sports Medicine 19:

Ren K, Dubner R 1996 Enhanced descending modulation of nociception

Waddell G, Newton M, Henderson I, Somerville D, Main C J 1993

A fear-avoidance beliefs questionnaire (FABQ) and the role of fear­ avoidance beliefs in chronic low back pain and disability. Pain 52: 157-168

Ren K, Wei F, Dubner R, Murphy A, Hoffman G E 2000 Progesterone attenuates persistent inflammatory hyperalgesia in female rats: 865: 272-277 113-118 involvement of spinal NMDA receptor mechanisms. Brain Research,

Wei F, Ren K, Dubner R 1998 Inflammation-induced Fos protein

expression in the rat spinal cord is enhanced following dorsolateral or ventrolateral funiculus lesions. Brain Research 782: 136--1 41

Sandki.ihler J 2000 Learning and memory in pain pathways. Pain 88:

Wei F, Dubner R, Ren K 1 999a Nucleus reticularis gigantocellularis and nucleus raphe magnus in the brain stem exert opposite effects on

N e u roph ysio l ogy of p a i n a n d p a i n m o d u l ation


behavioural hyperalgesia and spinal Fos protein expression after Wei F, Dubner R, Ren K 1999b Dorsolateral funiculus-lesions wunask peripheral inflammation. Pain 80: 127-141

Woolf C J, Doubell T P 1994 The pathophysiology of chronic pain: increased sensitivity to low threshold A�-fibre inputs. Current Opinion in Neurobiology 4: 525-534

inhibitory or disfacilitatory mechanisms which modulate the effects of innocuous mechanical stimulation on spinal Fos expression after inflammation. Brain Research 820: 112-116

Woolf C J, Mannion R J 1 999 Neuropathic pain: aetiology, symptoms, Woolf C J, Salter M W 2000 Neuronal plasticity: increasing the gain in Woolf C J, Shortland P, Coggeshall R E 1 992 Peripheral nerve injury pain. Science 288: 1 765-1768 mechanisms, and management: Lancet 353: 1959-1964

Wilcox G L 1991 Excitatory neurotransmitters and pain. In: Bond M, Woolf C J, Charlton J E (eds) Pain research and clinical Elsevier, Amsterdam, pp 97-117 management. Proceedings of the VIth World Congress on Pain.

Willis W D 1985 Anatomy and physiology of nociceptive ascending pathways. Philosophical Transactions of the Royal Society of London Series B 308: 253-268

Woolf C J, Safieh-Garabedian B, Ma Q-P, Crilly P, Winter J 1994 Nerve hypersensitivity. Neuroscience 62: 327-331

triggers central sprouting of myelinated afferents. Nature 355: 75-78 growth factor contributes to the generation of inflammatory sensory

Woolf C J 1983 Evidence for a central component of post-injury pain Woolf C J 1984 Long term alteration in the excitability of the flexion decerebrate rat. Pain 18: 325-343 reflex produced by peripheral tissue injury in the chronic hypersensitivity. Nature 306: 686-688

Wright A, Thurnwald P, O'Callaghan J, Smith J, Vixenzino B 1994 Pain 2: 83-97

HyperalgeSia in tennis elbow patients. Journal of Musculoskeletal

Zhuo M, Gebhart G F 1990 Spinal cholinergic and monoaminergic

receptors mediate descending inhibition from the nuclei reticularis gigantocellularis and gigantocellularis pars alpha in the rat. Brain Research 535: 67-78

Woolf C J 1994 A new strategy for the treatment of inflammatory pain: Woolf C J, Costigan M 1999 Transcriptional and posttranslational prevention or elimination of central sensitisation. Drugs 47: 1-9

Zhuo M, Gebhart G F 2002 Modulation of noxious and non-noxious the rat. Journal of Neurophysiology 88: 2928-2941

plasticity and the generation of inflammatory pain. Proceedings of 96: 7723-7730 the National Academy of Sciences of the United States of America

spinal mechanical transmission from the rostral medial medulla in





The effect of pain on motor control
M. Galea

CHAPTER CONTENTS Introduction pain
174 174 173

Differences between cutaneous and deep Sites uf termination of nociceptive afferents in the dorsal horn Referral of pain Muscle pain Joint pain
174 175 175 174

Pain and the brain pain

Changes in reflex activity in response to Stretch reflex Group
176 176 176 176

Tension feedback reflex

II reflexes

Flexor reflex

Reflex changes following painful stimuli Withdrawal reflexes Stretch reflexes
177 176

Abnormal muscle activity in chronic pain Muscle hyperactivity Muscle inhibition Wider effects Reflex model Systems model
179 179 179 179 181 180 177 178


Models of motor control Hierarchical model

Nervous system changes

A dynamical view of pain and motor control


Pain involves a complex series of sensory and behavioural responses. Under normal circumstances, pain is the conse­ quence of stimuli that either threaten or cause injury (Willis 1989). The responses to such stimuli include not only the perception of pain which may have a range of sensory qual­ ities, but also responses such as arousal, distress, somatic and autonomic reflexes and endocrine changes. Such stim­ uli activate a sequence of events involving nociceptors, ascending somatosensory pathways, the thalamus and the cerebral cortex. Motivational-affective responses are trig­ gered by another, related, system that operates in parallel with this sensory-discriminative system ( Melzack & Casey 1968). Even the memory of pain can condition behaviour (fear avoidance). Musculoskeletal pain influences motor performance and because of this, the behaviour of a person in pain is an important element of an assessment by a health professional. Since the biological role of pain has been regarded as a protective one, the commonly observed lack of mobility, choice of posture and/or avoidance of movement of the person in pain have been interpreted as protective mechanisms that avoid further injury. While this is a logical response of the human system to acute pain, it does not serve a useful purpose in chronic pain situations and does not provide a framework for rehabilitation. Despite the high prevalence of pain arising from deep structures such as muscle and articular tissues, much of our knowledge about mechanisms of acute and chronic pain has been derived from studies of the effects of nociceptive cutaneous stimuli in experimental animals, experimental subjects and people in pain. However, although it might appear reasonable to generalize these findings to pain pro­ duced by stimulation of deep tissues ( muscle, joints and viscera), clinical studies suggest that there may be impor­ tant differences between cutaneous and deep pain. This review will focus on the effects of deep pain on motor con­ trol. While there have been a large number of studies that have investigated changes in reflexes in response to a painful stimulus, more recent work has identified changes



in the patterns of muscle activation in response to pain. These findings will be discussed with reference to motor control theories. One of the problems in using painful stimuli in the experimental situation is stimulus control. Unlike stimuli used for the study of other sensory systems which can be well-defined and repeatedly applied, painful stimuli are difficult to control, and their repeated application can dra­ matically alter the behaviour of the sensory receptors (Willis 1989). There are also difficulties in interpretation of studies of motor behaviour of subjects with clinical pain disorders because of the variety of conditions and hetero­ geneity within these populations. In order to investigate the mechanisms underlying changes in motor performance, a number of experimental models in both animals and humans have been developed. Intramuscular injections of hypertonic ( 5%) saline, introduced in the 1930s by Kellgren (1938), are being used more commonly to induce pain in previously pain-free subjects. Muscle pain induced by hypertonic saline appears to have very similar characteris­ tics to the subjectively perceived qualities and the motor performance effects of clinical musculoskeletal pain (Arendt-Nielsen et aI1996). The muscles selected for injec­ tion are those usually affected in clinical conditions. Thus, injections have been made into paraspinal muscles and muscles of the jaw, wrist and lower leg. Animal models have been developed for acute joint inflammation or arthri­ tis involving the injection of kaolin (aluminium silicate) and carrageenan (a sulphated polysaccharide) into a joint. The injection of complete Freund's Adjuvant into a joint pro­ vides a model for a chronic inflammatory condition. In both models, joint swelling and increased temperature are observed, along with behavioural changes such as limping, guarding of the affected limb and hyperalgesia to heat and mechanical stimuli. Such experimental models of joint pain cannot be used in humans for obvious reasons, although some researchers have injected saline into the knee joint to cause an artificial effusion ( Hopkins et al 2000). Investigations of the effects of joint pain in humans have focused on clinical subjects with acute joint sprains or chronic joint conditions.

somatic and visceral inputs. This multisegrnental distribu­ tion may account for the diffuse and poorly localized nature of deep pain sensations. Cutaneous and deep affer­ ents involved in nociception also differ in the laminar dis­ tribution of their terminals in the spinal cord. The superficial dorsal hom (laminae I and II) is the site of ter­ mination of myelinated and unmyelinated cutaneous noci­ ceptive afferents, whereas deep afferents (from muscles and joints) involved in nociception terminate in lamina I and/ or in laminae IV and V ( Mense 1986, Craig et a11988, Hoheisel et a11989, Yu & Mense 1990). Nociceptive-specific neurons are predominantly in lamina I and the outer part of lamina II, whereas wide-dynamic range cells are more concen­ trated in lamina V. These differences in terminal projections imply that there are different intraspinal connections and therefore differential central processing of cutaneous and deep inputs occurs in the dorsal hom.
Referral of pain

DIFFERENCES BETWEEN CUTANEOUS AND DEEP PAIN Sites of termination of nociceptive afferents in the dorsal horn

Mechanosensitive receptive fields (RF) of muscle nocicep­ tors extend over a small portion of the muscle (Mense & Meyer 1985). At the level of dorsal hom, RFs of neurons processing information from muscle nociceptors are also small, but many of the cells have multiple RFs from deep tissues and often additional input from the skin. The multi­ plicity of RFs and the convergence from skin and deep tis­ sues provide support for the convergence-projection theory (Ruch 1946) and may explain the referral of deep pain to cutaneous regions as well as the diffuse nature of deep pain sensations. This is exemplified by the results of an experiment in which hypertonic saline was injected into masseter muscle. There was a loss of mechanosensitivity to threshold level monofilament stimuli applied to facial skin, not only at the site of pain but also on the contralateral side (Stohler et al 2001). This suggests that muscle nociceptors excite neurons in the trigeminal nucleus caudalis that sup­ press thalamic transmission from touch receptors on both sides of the face. The number of dorsal hom cells that receive information exclusively from muscle nociceptors appears to be relatively small. The more medially a cell is located in the dorsal hom the more distal is the site of its deep receptive field (Yu & Mense 1990). This somatotopic arrangement may be important for the control of local reflexes.
Muscle pain

Cutaneous nociceptive afferents have well-circumscribed termination sites in the dorsal hom, typically up to 500 microns in the rostrocaudal direction. Deep nociceptive afferents, on the other hand, especially those innervating visceral tissues, have an extensive longitudinal distribution over several segments in the dorsal hom (Sugiura et al 1989). Dorsal hom neurons responsive to these inputs often have very large receptive fields and receive convergent

Nociceptors are found throughout skeletal muscle, most densely in the region of tendons, fascia and aponeuroses (Stacey 1969). There is a view that muscle pain may become chronic through a series of vicious cycles. Lesions of muscle are likely to induce the release of endogenous sensitizing and pain-producing substances such as kinins and prostaglandins. These substances cause vasodilation and may cause local oedema in high concentrafions. The

The effect of pain on motor control


increase in interstitial pressure may compress veins, lead­ ing to venous congestion and ischaemia. Ischaemia, in turn, is a powerful promoting factor for the release of noci­ ceptive substances such as bradykinin or prostaglandin E2. These have a sensitizing effect on nociceptors. Strong input from muscle nociceptors can lead to increases in the excitability of dorsal horn neurons and a contribution to central sensitization, forming a vicious circle ( Mense 1991). Ischaemia may lead to failure of the calcium pump and local contracture, which may impair local circulation and enhance the ischaemia. Such a mechanism has been pro­ posed for the development and maintenance of trigger points (Simons 1990). Deep somatic pain has been associated with local increases of muscle tone (Travell & Simons 1983). The term 'vicious cycle' has also been given to a process whereby nociceptive afferent fibres stimulated by a painful lesion activate gamma motoneurons, which increases the dis­ charge rate of the muscle spindles. This leads to activation of the muscle via monosynaptic connections with the alpha motoneuron and hypertonus in the affected muscles Gohansson & Sojka 1991). This in turn leads to ischaemia and the processes described above. This model is discussed further below.
Joint pain

Nociceptors in joints are located in the joint capsule and lig­ aments, bone, periosteum, articular fat pads and around blood vessels, but not in the joint cartilage (Wyke 1981). In conditions such as osteoarthritis, joint pain is the major symptom, with movement related pain the most common type of pain reported. The stimulus for pain in the chronic stages is most probably mechanical, since the anatomy of the joint is abnormal and leads to mechanical stresses on the capsule, ligaments and peri-articular tissues. Some inflammation may occur, leading to the release of chemical stimuli. Weakness has been observed in muscles surround­ ing a painful joint, although it is not clear that pain is the stimulus for this (Shakespeare et al 1985, Fahrer et al 1988, Fischer-Rasmussen et a12001). Increased background activity as well as increased responses to noxious and innocuous joint movement in A�, Ao and C afferent fibres have been observed following acute inflammation. 'Silent nociceptors' are neurons that do not respond to peripheral mechanical stimuli in normal intact tissue, but begin to respond to innocuous and nox­ ious stimuli, as well as to pressure and joint movement fol­ lowing joint inflammation (Schaible & Schmidt, 1988).

A matrix of structures in the nervous system has been iden­ tified as being variably activated during pain experience. The type, duration and location of stimulus differs consid­ erably between clinical studies (neuropathic pain, idio-

pathic pain, cancer pain and post-stroke pain have all been studied). Experimental stimuli include ethanol injection, hot and cold probes, and each pain stimulus has varied with respect to intensity and quality. The thalamus represents the final link in the transmis­ sion of impulses to the cerebral cortex, processing almost all sensory and motor information prior to its transfer to corti­ cal areas. There is a differential pattern of connectivity of afferent information in the thalamus, with spinothalamic afferents mediating the sensory-discriminative aspects of pain terminating in the ventral posterolateral nucleus, and medial thalamic nuclei ( the central lateral nucleus, the intralaminar complex and the mediodorsal nucleus), receiving additional information from the reticular forma­ tion, the cerebellum, and globus pallidus (see Galea 2002 for review). The diffuse projections of the intralaminar nuclei to many different areas of the cortex have been con­ sidered to be part of a non-specific arousal system, but it is also possible that their role is concerned with affective states induced by a painful stimulus (see Galea 2002 for review). The basal ganglia are associated with planned action (Brooks et al 1993) and movement ( Colebatch et al 1991). Their connections through the thalamus with the pre­ frontal cortex, supplementary motor cortex, motor cortex and anterior congulate cortex (Cote & Crutcher 1991) form a circuit associated with motor preparation or response selection. A number of brain stem structures, including the peri-aqueductal grey matter (Bernard & Bandler 1998) and the reticular formation, have extensive connections with all levels of the nervous system and are involved in the noci­ ceptive, autonomic and motor systems as well as in pain modulation. Multiple cortical areas are activated by painful stimuli, including the primary somatosensory cortex (Bushnell et al 1999), secondary somatosensory cortex, anterior cingulate cortex (Talbot et al 1991), insula, prefrontal cortex (Treede et al 1999) and supplementary motor area (Coghill et al 1994). These cortical regions also give rise to corticospinal projection (Galea & Darian-Smith 1994). This distributed activation of cerebral structures reflects the complex nature of pain, involving discriminative, affective, autonomic and motor components (Coghill et al 1994). Parietal areas are mainly concerned with the sensory-discriminative aspects whereas frontal-limbic connections subserve the affective dimension of pain experience. Corticospinal projections are the only direct link between the sensorimotor cortex and the spinal cord and form a par­ allel, distributed system arising from cortical areas with com­ plex cortical and thalamic interconnections, and converging on different parts of the spinal circuitry. There are projections to all laminae of the dorsal horn, mainly from postcentral cortical areas, including laminae containing spinothalamic neurons (Coulter & Jones 1977, Cheema et al 1984, Ralston & Ralston 1985). Precentral cortical areas project predomi­ nantly to laminae vn and VITI, with neurons from primary



motor cortex projecting directly onto motoneurons in lamina IX (Galea & Darian-Smith 1997, Maier et al 1997). The dorsal caudal cingulate area projects to the dorsal portion of the intermediate zone of the spinal cord (Dum & Strick 1996). It must be emphasized that apart from the direct projections to motoneurons, the majority of these corticospinal projections terminate on intemeurons that are also part of spinal circuits involved in movement. The 'state' of the intemeurons is dependent on the combined influence on descending path­ ways, spinal interactions and afferent input and is related to the upcoming motor task. It is well known that stimulation of corticospinal projec­ tions from primary and secondary somatosensory areas can result in primary afferent depolarization (PAD) in fibres of the dorsal root (Carpenter et al 1963, Andersen et al 1964). Stimulation of sensorimotor cortex can elicit both excitatory and inhibitory responses in dorsal hom neurons, particu­ larly in laminae IV and V (Lundberg et al 1962, Wall 1967, Fetz 1968). Corticospinal projections to the superficial lam­ inae from primary somatosensory cortex may directly mod­ ulate nociceptive-specific neurons (Cheema et al 1984). Posterior parietal cortex has connections with the primary somatosensory cortex and other polymodal association areas, including the limbic system (Cavada & Goldman­ Rakic 1989) and is part of a general attentional system. The cingulate cortex is involved in affective and motor behav­ iour ( Devinsky et al 1995). This region contains neurons that fire in anticipation of pain and could therefore be involved in avoidance behaviour (Koyama et al 1998). The insula receives converging information about all five sen­ sory modalities and has extensive connections with the lim­ bic system and the spinal cord (see Galea 2002 for review). The role of many of these areas in the control of movement is still under investigation. However all these regions, through the corticospinal tract, may exert a modulatory effect on both motor and sensory functions, including pain.

Stretch reflex

The largest diameter sensory nerves, the Ia fibres from muscle spindles, make monosynaptic excitatory synapses on their own motoneurons and disynaptic inhibitory synapses onto antagonist motoneurons. This reflex is induced by stretching of muscle spindles and results in con­ traction of the stretched muscle and reciprocal inhibition of the antagonist. Muscle spindles themselves are innervated by gamma motoneurons, which regulate the sensitivity of the muscle spindle to stretch ( Fig. 13.1A).
Tension feedback reflex

The large diameter group II fibres from the Golgi tendon organs (GTOs) make disynaptic inhibitory connections �ith �otoneurons and excitatory connections with antago­ rust (mverse myotatic reflex). The effects on the motoneu­ rons from a given muscle are the reverse of those in the stretch reflex. The GTOs are especially sensitive to tension arising from muscle contraction. The effect of a muscle con­ traction is to decrease the amount of contraction of that muscle but increase the excitation of opposing muscles. Combined with the stretch reflex, the inverse myotatic reflex contributes to overall muscle stiffness. The y motor innervation of spindles keeps the muscle under resting ten­ sion and the GTO becomes exquisitely sensitive to tension changes due to active muscle contraction (Shepherd 1994) (Fig. 13.18).
Group II reflexes

The group II afferents from muscle spindles arise from the group II (flower spray) endings in chain fibres and make disynaptic connections onto motoneurons. Excitatory con­ nections are directed mainly to flexor muscles, and inhibitory connections to extensor muscles ( Fig. 13.1C).
Flexor reflex

Measures of reflex activity ( H reflex, nociceptive reflex and blink reflex) have been used in human subjects as an index of subjective pain, as they correlate with other physiologi­ cal parameters and with verbal report (Gracely 1994). However, changes in reflex activity have been reported in relation to muscle or joint pain. Reflexes are important ele­ ments of motor activity, first identified and categorized in the 19th century by Marshall Hall (1790-1857), but described clearly as a structural and functional entity by Sherrington and co-workers in the early part of the 20th century. Sherrington's view of the reflex as an elementary unit of behaviour dominated the field of motor control until recently. A brief review of the best known reflexes mediated by the spinal cord, and categorized in terms of their main sensory input, provides a useful framework for later dis­ cussion of motor control theories.

A noxious stimulus applied to skin or muscle characteristi­ cally produces withdrawal of the affected limb. This is termed the flexor reflex, and can be mediated by a wide range of receptors collectively referred to as the flexor reflex afferents ( FRAs) ( Eccles & Lundberg 1959). Where Sherrington (1910) described the flexion reflex as a mecha­ nism for withdrawing a limb from noxious stimuli, Lundberg promoted the concept that the FRA systems are used during normal movements (Lundberg et al 1987) (Fig. 13.1D).
Reflex changes following painful stimuli Withdrawal reflexes

Noxious stimulation of articular or muscle tissues can evoke activation of a limb flexion reflex (Gardner 1950,

The effect of pain on motor contro l



have been viewed as responses that serve to protect the limb from further noxious stimulation, and as mechanisms counteracting excessive movement so as to prevent further damage to the joint or muscle tissues (Schaible & Schmidt 1985, Mense 1986, He et al 1988). Although painful stimuli may evoke flexion, painful stimuli are not necessary for segmental activation of FRA pathways. Nociceptive affer­ ent input presumably contacts spinal circuits involved both in flexion reflexes and perception of pain. Indeed, there is evidence for withdrawal reflexes in response to pain that are distinct from spinal motor pathways producing flexion (Schouenberg & Sj6lund 1983, McCrea 1994). High intensity stimulation of limb muscle afferents can produce a pro­ longed facilitation of the flexor reflex (Wall & Woolf 1984) and this effect may be related to pathophysiological responses to injury or inflammation.
Stretch reflexes



Cutaneous afferents i .J-----tci

There is conflicting evidence about the influence of pain on stretch reflexes. Activation of muscle nociceptors leads to increased fusimotor firing and increased sensitivity of the Ia muscle spindle afferents to stretch ( Appelberg et a11983, Johansson et a11993, Pedersen et a11997, Wang et al 2000). In contrast, Mense & Skeppar (1991) demonstrated inhibi­ tion of extensor gamma motoneurons following induced inflammatory muscle pain in the cat. There have been equivocal results in human experimental subjects. Matre et al (1998) demonstrated facilitation of the stretch reflex in the soleus and tibialis anterior muscles following injection of hypertonic saline, although without a corresponding increase in the amplitude of the H reflex, indicating that the excitability of the alpha motoneuron pool was unchanged. This increase in the stretch reflex disappeared with volun­ tary contraction of the muscle under investigation. On the other hand, Zedka et al (1999) found no increase in reflex activity in erector spinae muscles injected with hypertonic saline. These discrepancies could be partly due to differ­ ences in the techniques used to elicit reflex responses as well as differences in the function of the experimental mus­ cle selected.

Figure 1 3 . 1 Neural circuits for the main types of spinal reflexes. Inhibitory terminals and interneurons are filled (black), while exci­ tatory connections are unfilled (white). A: Stretch reflex - detects phasic stretch of muscle and contributes to the control of move­ ment. B: Inverse myotatic reflex - detects tension and contributes to the control of muscle force and stiffness. C: Group II reflex detects steady stretch of muscle and contributes to postural con­ trol. D: Flexor reflex - detects harmful stimuli and serves to with­ draw a limb from harm.


Painta11961, Mense 1986, He et aI1988). Injection of algesic chemicals into the temporomandibular joint results in a sustained reflex increase in activity in tongue and jaw­ opening muscles; weaker excitatory effects are observed in jaw-closing muscles (Broton & Sessle 1988). These effects

Observations of increased muscle tone associated with painful muscles (Travell & Simons 1983) led to attempts to explain these findings. One proposal is the facilitation of the gamma motor system by muscle pain (Johansson & Sojka 1991). The gamma motoneurons receive information from a wide variety of inputs, including the skin and joint ligaments, which contributes to the coordination of muscle tone, posture and movement. Group III and IV muscle afferents (comprising mechanoreceptors and nociceptors) are known to have a powerful influence on gamma motoneurons (Johansson et al 1989, Mense & Skeppar



1991). The hypothesis proposed by Johansson & Sojka involves stimulation of group III and IV nociceptors by chemical inflammatory mediators released by muscle con­ traction. Metabolites released during muscle contraction have been shown to activate group III and IV receptors (Rybicki et al 1985). These nociceptors synapse with and excite gamma motoneurons that stimulate muscle spindles, causing an increase in the output of the group Ia and II afferents. This stimulates alpha motoneurons, causing fur­ ther muscle contractions that generate further metabolites and complete a positive feedback loop. Increased activity in the group II afferents excites the gamma motoneurons, which stimulate the muscle spindle. This constitutes a sec­ ond positive feedback loop that can be maintained without group III or IV nociceptive input (Gladden et aI1998). There is evidence that increased concentrations of metabolites from muscle contractions, including lactic acid, potassium chloride Govanovic et aI1990), arachidonic acid (Djupsjobacka et al 1994), bradykinin (Djupsjobacka et al 1995, Pedersen et al 1997) and 5-HT (Djupsjobacka et al 1995), result in increases in gamma motoneuron activity and sensitivity of the muscle spindle afferents, thereby leading to increases in muscle stiffness. Increased gamma motoneuron activity has been induced by fatiguing con­ tractions ( Nelson & Hutton 1985, Ljubisavljevic & Anastasijevic 1994). Experimentally induced pain has been shown to be associated with changes in the fusimotor sys­ tem (Thunberg et aI2002). Neck and paraspinal muscles are rich in muscle spindles (Richmond & Abrahams 1975, Amonoo-Kuofi 1982, Boyd-Clark et al 2002). These spin­ dles, many of which lack a bagl fibre, are controlled by the static fusimotor system and concerned mainly with pos­ tural muscle activity (Price & Dutia 1989). Thus changes in muscle stiffness in these muscles are likely to cause distur­ bances in motor coordination and proprioception. Indeed, stimulation of fusimotor activity through application of muscle vibration to neck muscles has been shown to cause motor and balance disturbances ( Lund 1980, Biguer et al 1988).
Muscle inhibition

Despite this experimental support for the Johansson-Sojka hypothesis, clinical observations suggest that there is inhi­ bition of muscle activity during muscle pain. An alternative model, the pain-adaptation model, proposes that muscle dysfunction is a normal protective adaptation and is not one of the causes of pain (Lund et al 1993). Nociceptive input from painful muscle, joint and skin converge on interneurons at the segmental level and as a consequence motoneurons to the painful muscles are inhibited. In this way the amplitude of movement will be limited and poten­ tially prevent further damage. The relationship between muscle pain and muscle activ­ ity has been examined at rest, during static contractions and during dynamic tasks (Graven-Nielsen et aI1997). A review

of this issue is complicated by different experimental para­ digms and subject types. EMG activity has been shown to be higher at rest in some studies of subjects with induced muscle pain (Cobb et al 1975) but studies of patients in pain, for example low back pain (Nouwen & Bush 1984, Sherman 1985), myofascial pain (Durette et al 1991) or experimentally induced muscle pain (Stohler et al 1996), have shown no such increased activity. The maximum voluntary contraction in painful muscles has been shown to be reduced in experimentally induced pain (Graven-Nielsen et al 1997) and in a number of disor­ ders, including temporomandibular joint dysfunction ( Molin 1972), fibromyalgia Gacobsen & Danneskiold­ Samsoe 1987, Backman et al 1988) and low back pain (Thorstensson & Arvidson 1982, Kankaanpaa et aI1998). Studies that have investigated the effect of pain on movement have shown that there is an inhibition of the painful muscle and facilitation of its antagonist. In activities such as walking there are changes in the coordination of muscle activity leading to reductions in movement ampli­ tude and reduced stride time following experimentally induced muscle pain in muscles of the lower leg (Graven­ Nielsen et a11997, Madeleine et aI1999a). Zedka et al (1999) showed that injections of hypertonic saline into the erector spinae on one side resulted in a reduction in the velocity and range of voluntary trunk motion and a reduction in EMG amplitude in the affected muscle, consistent with the pain-adaptation model of Lund et al (1993). When subjects voluntarily overcame this guarding strategy and produced identical trunk movements before and during pain, the reduced amplitude in EMG activity persisted, indicating that the observed changes involve more than just a strategy to reduce movement. In contrast, no pain-induced changes in motor unit activity have been observed in cases of exper­ imentally induced pain in extensor carpi ulnaris (Birch et al 2000). Changes in muscle activity and coordination during muscle pain therefore appear to be dependent on the func­ tional role of the muscle and the level of muscle activity. These experimental manipulations can only provide an indication of the effect of acute pain. Other investigations have identified specific deficits in the chronic pain situa­ tion. Patients with chronic low back pain fatigue faster (Kankaanpaa et a11998) and have poorer balance perform­ ance and delayed postural response times compared with healthy control subjects ( Radebold et al 2000, 2001, Newcomer et al 2002). Investigations of patients with chronic low back pain have identified a deficit in the recruitment of the transversus abdominis during a postural perturbation produced by rapid arm movement (Hodges & Richardson 1996) or movement of the lower limb (Hodges & Richardson 1998), resulting in sub-optimal control of the lumbar spine in preparation for movement. A study in nor­ mal subjects has shown that trunk muscle fatigue is one fac­ tor that can alter anticipatory postural adjustments (Allison & Henry 2002) and there is individual variability in the preparatory strategies used to deal with sudden trunk load-

The effect of pain on motor control


ing (Lavender et al 1993). Such impairments of postural control could compromise the stability of the lumbar spine, making a person vulnerable to further injury. A similar phe­ nomenon has been described in patients with chronic wrist pain. Such patients demonstrate a disturbance of fine motor control of the wrist on the unaffected side and it has been argued that this incoordination might result in additional overuse injury (Smeulders et al 2002). Wasting of another muscle contributing to stability of the lumbar spine, multifidus, has been reported in patients with low back pain, with the site of wasting corresponding to the clinically determined level of symptoms ( Hides et al 1994), suggesting a localized reflex inhibition. Reflex inhibi­ tion of the quadriceps has also been associated with knee joint pathology, especially effusion (Shakespeare et al 1985, Young et al 1987), and patients with patellofemoral pain are reported to have deficits in the timing of activation of the vastus medialis during a functional stepping task ( Cowan et al 2001). While these findings are consistent with the pain-adaptation model, the resultant motor deficit not only limits the amount of movement, but also makes the affected region vulnerable to further damage. However, it is not known whether these motor deficits are compensatory mechanisms developed in response to pain or whether they are, in fact, predisposing factors to injury.
Wider effects

theories provides a context for drawing together some of these findings.

Motor control models and theories have generally been developed around the following questions: 1. What is the basic unit of nervous system organization in relation to the basic unit of motor function? 2. What principles apply to the organization of motor control?
Reflex model

It also needs to be highlighted that nociceptive stimuli can have effects that extend beyond the local region. Nociceptive stimuli applied to extensor digitorum brevis in the foot produces a depression of Ia excitation and Ib inhi­ bition of the soleus motoneurons through intercalated interneurons (Rossi et al 1999). Conversely, facilitation of soleus motoneurons has been observed following the induction of an artificial knee effusion, which usually inhibits the quadriceps ( Hopkins et al 2000). Injection of hypertonic saline into the trapezius muscle on one side prior to the performance of a low-load, repetitive work task resulted not only in a reduction of muscle activity in the injected muscle but also a reduced working rhythm, a ten­ dency to increase the amplitude of arm movements and a prolongation of the duration of the role of the non-affected arm in the task (Madeleine et al 1999b). Changes in rhyth­ mic activity, such as chewing (Westberg et al 1997) or loco­ motion ( Martin & Arendt-Nielsen 2000), following hypertonic saline into the masseter or soleus respectively, demonstrate that the performance of other muscle groups participating in the motor patterns is also affected, and possibly reflecting a change in the underlying motor programmes. These studies have generated much data; however, there is still no clear consensus as to the effect of pain on motor control. This is perhaps partly because of the ten­ dency of many of these studies to consider only local stim­ ulus-response issues. An examination of motor control

The reflex model of motor control originated with Sherrington (1947) who found that specific stimuli such as stretch or pain induced distinct stereotyped movements called reflexes. An underlying assumption of this model is that afferent input is a prerequisite for motor output. The concept of feedback in relation to motor control was intro­ duced much later with the view that the human motor system might be controlled in ways similar to the control of mechanical systems. One such system is a servo­ mechanism that relies on feedback signals to maintain a constant output. Building on Sherrington's (1947) original notion of reflexes as elementary units of motor behaviour (see Fig. 13.1), successive researchers have described the circuitry by which reflex pathways were coordinated. Lundberg (1979) showed that different reflex pathways may share common interneurons. One of the best examples of this principle is the Ia interneuron. This neuron not only mediates inhibi­ tion of antagonistic muscles in the stretch reflex, but is also part of an inhibitory pathway from flexor reflex afferents onto motoneurons, and a nodal point for control of spinal neurons by descending pathways such as the corticospinal, rubrospinal and vestibulospinal tracts which are important for skilled movement and postural tasks ( Fig. 13.2).
Hierarchical model

The concept of the nervous system being organized as a hierarchy has a long history in clinical neurology. For exam­ ple, we distinguish between the lower motoneuron in the spinal cord and the upper motoneuron in the brain stem or cerebral cortex. The neurologist Hughlings Jackson (1835-1911) formulated the idea that there were successive levels of motor control in the nervous system, with the con­ trol of automatic or reflexive movements by lower levels and purposive movements by higher levels ( Fig. 13.3). It is now realized that motor control in mammals is not strictly hierarchical and the historical distinction between volun­ tary and reflex control is becoming increasingly blurred. For example, every voluntary movement is associated with automatic postural adjustments that occur unconsciously.




Vestibulospinal tract

Conlricospinal Vestibulospinal tract tract Rubrospinal tract Propriospinal syslems

Cutaneous afferents FRA

FRA ---

Renshaw cells

Figure 1 3.2 Diagram illustrating the la inhibitory interneuron (la IN) as an integrating node in a number of spinal circuits. Excitatory neurons and synaptic terminals are unfilled (White), inhibitory con­ nections are filled (black). Reproduced from Shepherd 1998 with permission of Oxford University Press Inc. Key: FRA flexor reflex afferents; MN motoneuron; la afferent, sensory nerve from muscle spindle.
= =

Many volitional actions are often adjusted automatically by sensory feedback. There are two basic systems for control of movement: feedback or closed-loop systems and feedforward or open­ loop systems. Feedback control is usually required for slow movements or those requiring accuracy and this type of

Brain Central feedback

control is typical during the early phases of skill acquisi­ tion. The control of movement fluctuates between feedback and feedforward modes of control. Rapid movements and well-learned movements are performed with feedforward control. Sensory information cannot be used during rapid movements because the movement occurs faster than the nervous system can process the sensory information. The motor programme concept is based on an open-loop system of control and assumes that all movements are pre­ planned and stored in memory until required for action. Motor programmes have been defined as sets of muscle commands that are structured before a movement sequence begins, and that allow the entire movement sequence to be performed without the influence of periph­ eral feedback. Such programmes contain details of the order and timing of events and the relative force to be exerted. An example of such programmes is the ability of the spinal cord to generate intrinsic rhythms or repeating patterns of muscle activity. Studies of both vertebrates and invertebrates have demonstrated that this property lies within neural circuits forming central pattern generators. Central pattern generators are recognized as key organiz­ ing principles for understanding the mechanisms of loco­ motion, and other rhythmical activities such as respiration. These circuits generate essential rhythmical features of the motor pattern and receive sensory feedback signals (see Fig. 13.3). However, the linear view that underlies the idea of a closed loop system, inputs, outputs, stimuli and responses, with feedback closing the loop, is only useful for simple systems. In a more complex system, the concept of feedback is inadequate. Feedforward control models involve the delivery of information to other parts of the system to 'prepare' it for upcoming motor commands. Postural adjustments are an example of feedforward mechanisms and have been shown to precede self-initiated arm, leg and trunk movements to minimize postural instability that would otherwise have resulted (Lee 1980, Massion et al 1982, Frank & Earl, 1990). This type of control is dependent on the nervous system having an accurate internal model of the body and the external environment. An inappropriate internal model can lead to poor predictions about the sensory consequences of situations and actions, resulting in anticipatory movements that are ineffective or destabilizing (Horak 1991).
Systems model

Spinal cord Feedback Reflex feedback Muscles Motor output Environment from environment

Figure 1 3.3 Connections of spinal circuits. Key: CPG central pattern generator.

It is not reasonable to assume that each individual neural and muscular component of the human body is controlled separately by the nervous system, as this would be a major computational task. Bernstein (1967) proposed a solution that the multiple degrees of freedom for movement are con­ strained to act as synergies or coordinative structures (i.e. functional groups of muscles and joints that are constrained to act as a unit). Examples of these synergies are the pos­ tural movement strategies described by Horak &- Nashner

The effect of pain on motor control

18 1

(1986), which are used to recover stability in response to brief perturbations of the supporting surface (Fig. 13.4). In this view, control is exerted not over muscles or sensory receptors ( reflex model) nor over muscle activation patterns (hierarchical model) but over abstract aspects of motor behaviour, such as the relations between kinematic vari­ ables and the accomplishment of task goals. The dynamical action theory, an elaboration of systems theory, proposes that movement emerges naturally out of the complex interactions among many interconnected ele­ ments (physical, environmental and neural), without spe­ cific commands or motor programmes in the central nervous system. At the core of the dynamical action theory is the notion that human behaviour is governed by a generic process of self-organization, which refers to the spontaneous forma­ tion of patterns and pattern change. This idea takes into account the coordinative relations of various parts of a sys­ tem, as well as the environment in which the interaction between the parts takes place. The coordinative structures are the units of action and are functionally linked, but not necessarily mechanically linked ( Kelso & Tuller 1984). This view of motor control, that is that musculoskeletal variables are not controlled individually but are partitioned into a

smaller number of coordinative structures, has led to hypotheses about how these coordinative structures operate using the framework of task dynamics (Saltzman & Kelso, 1987). The critical issues are that the coordinative structures are constrained by the particular tasks that are being per­ formed, and that the units of action are specified in dynamic terms rather than kinematic or muscular variables. Thus the task will determine many features of the action, such as movement trajectory for example. This might explain the conflicting observations on the effect of pain in different muscle groups that have different functional roles.
Nervous system changes



'" 0 u.

Abd Ham Quad

Gas! A Norma! Tib


It needs to be recognized that neural representations of body parts and movements are labile throughout life, changing according to the amount they are activated by peripheral inputs. This phenomenon has been most graph­ ically demonstrated by the experiments of Merzenich and colleagues (1983, 1984) in which the reorganized cortical representations of the hand in the primary somatosensory cortex were mapped in the monkey following peripheral nerve lesions or digit amputation ( Fig. 13.5A). These find­ ings, and others illustrating similar reorganization in other brain regions, indicate that topographic maps in the adult brain are not hard-wired, but can vary depending on spa­ tial shifts in the collective activity of neurons with experi­ ence, a reorganization which is not haphazard but context-dependent. The brain itself, therefore, can be viewed as a dynamical system ( Kelso 1995). Byl et al (1996) showed in monkeys that rapid, repetitive, highly stereo­ typic movements can actively degrade the cortical repre­ sentations of sensory information guiding fine motor hand movements. This 'dedifferentiation' of sensory feedback information from the hand led to focal dystonia and subse­ quent lack of use of the hand ( Fig. 13.5B). Similarly, altered use of a part of the body will change the cortical represen­ tation of that part in sensorimotor cortical areas, and may therefore lead to prolonged movement dysfunction.
A dynamical view of pain and motor control








Abd Ha



� �



Studying individual elements of movement has increased our knowledge of the component parts of movement, but it has not contributed a great deal to our understanding of the func­ tion of individual elements related to behaviour as a whole. There is still a lack of understanding about how these ele­ ments function together or change with altered input, such as a painful stimulus. What is clear from this review is that:

Figure 13.4 Muscle synergy and body motion associated with the ankle strategy for controlling forward sway (A) and backward sway (B). Reproduced from Horak & Nashner 1986 with permission of the American Physiological Society.

groups of cells rather than single cells are the main units of activity in the nervous system functional synergies of muscles rather than single muscles are the main units of motor control and coordination of action there is distributed control of these systems.



A Location map 3b (SI proper)

o o


l mm

Multiple RFs: dorsaVglabrous surfaces

L-J �




Multiple RFs:

across adjacent digits

Representation order

C Normal map Dorsum



• Dorsal surface o Multiple RFs: dorsal/glabrous o Multiple RFs: adjacent digits
a = all digit segments p = proximal segment m = middle segment d

05 04 03 02 01

distal segment

05 04

02 01 o
Portion deprived by nerve section E Reorganization after nerve section Multiple RFs per cortical penetration

1 cm
Single RFs per cortical penetration


03 removed


Figure 13.5 Left-hand panel: Reorganization of primary somatosensory cortex (area 3b) in owl monkeys. A: Dorsolateral view of an owl monkey brain showing the location of the hand representation within area 3b. B: Ventral surface of the hand split along the palm to reflect the somatotopic pattern of hand representation. C: Pattern of hand representation in area 3b. D: Portion of the hand representation deprived of normal activation by section of the median nerve. E: Somatotopic pattern of the reorganized cortex months after median nerve section. Most of the activation is from receptors on the dorsum of the digits 1 3. F: Reorganization after D3 removal. Digits and pads of the hand are tradition­ ally numbered. Insular (I). hypothenar (H), and thenar (T) pads are indicated. Reproduced from Kaas 1992 with permission of MIT Press. Right-hand panel: Diagram illustrating an experience induced degradation of the hand representation in somatosensory cortex of a monkey following training on a hand-squeezing strategy. A, B, and C: Receptive fields on the hand and reconstruction of the topographic represen­ tations of the hand surface on the somatosensory cortex (area 3b) with penetration sites correlated with receptive fields on digits D and E, illustrating an abnormal 'map' of the hand. The magnitude of area 3b with multiple receptive fields was so extensive that separate drawings were necessary to clearly represent this (illustration A representing the overlap extending from the dorsal to the glabrous surface and illus­ tration B representing the overlap to adjacent digits). D and E: Outlines of territories over which neurons were driven with receptive fields, larger than normal size, on broad hand surfaces, with abbreviations of the cortical penetration sites matched to the topographical drawing in illustration C. Reproduced from Byl et al 1 99 7 with permission of the American Physical Therapy Association.

A reductionist approach investigating individual circuits is therefore unlikely to provide an indication of how the whole system operates, particularly with a multidimen­ sional phenomenon such as pain. Kelso (1995) has provided an example illustrating the key concepts of pattern formation that is useful for understand­ ing how different elements may work together to bring about a coordinated whole. The Rayleigh-Benard experi­ ment involves heating a fluid from below and cooling it from above. This is an open system, activated by the appli-

cation of a temperature gradient. The fluid contains many molecules and as temperature increases slightly, the heat is dissipated among them in a random fashion, and there is no visible motion of the fluid. As temperature increases, the fluid now begins to move as a coordinated whole in an orderly rolling motion called convection rolls (Fig. 1�.6A). The temperature gradient is the driving influence behind the motion, a so-called control parameter. However, this control parameter does not prescribe or contain the code for the emerging pattern. As Kelso (1995) states, rolling motions

The effect of pain on motor control




Fig ure 1 3 . 6 Convection patterns in flu id that is heated (refer to text for description). A reproduced from Kelso 1 995 with permission of MIT Press. B reproduced from Velarde Et Normand 1 980 with permission of Alan D. Isel in.

Actions are modulated according to changing environ­ mental circumstances, but there are 'rules', also known as constraints, that preserve qualitative aspects of a move­ ment's structure. One of these is that timing of activity in components of a functional unit is generally independent of the amplitude of the activity. Constancy in timing relation­ ships in muscle activity has been most clearly demon­ strated in studies of locomotion, for example the duration of the step cycle decreases when speed of locomotion increases, and also in other rhythmical activities such as mastication (Grillner 1975). Since pain appears to disrupt qualitative aspects of movement such as timing of muscle activity in certain tasks ( see, for example, Hodges & Richardson 1999, Martin & Arendt-Nielsen 2000), it is almost certainly a critical control parameter that can lead to radical alterations in the organization of motor activity and lead to dysfunctional movement patterns. Examples of this are the altered patterns of activity of the abdominal muscu­ lature in low back pain (Hodges & Richardson 1996, O'Sullivan et al 1997). Because of the adaptability of the nervous system, such patterns are likely to persist. This view of the motor control system provides a framework for understanding the effect of pain on motor activity, and a basis for the development of specific rehabilitation strate­ gies to reverse the dysfunction.

are not the only possibilities. In an open container surface tension may also affect the flow, and its net effect is tessela­ tion of the surface and the formation of hexagonal cells (Fig. 13.6B). Two quite distinct mechanisms can give rise to the same dynamic pattern, and conversely the same mechanism can give rise to different patterns, i.e. the mechanism­ pattern relationship is not fixed.
Allison G T, Henry S M 2002 The influence of fatigue on trunk muscle Biomechanics 17: 414-417 response to sudden arm movements: a pilot study. Clinical


muscle pain joint pain reflex activity

abnormal m uscle activity motor control models dynamical systems

Bernstein N A 1967 The coordination and regulation of movements. Biguer B, Donaldson M L, Hein A, Jeannerod M 1988 Neck muscle in man. Brain 111: 1405-1424 Pergamon Press, Elmsford, New York

Amonoo-Kuofi H S 1982 The number and distribution of muscle Anatomy 135: 585-599 spindles in human intrinsic postvertebral muscles. Journal of

vibration modifies the representation of visual motion and direction

Andersen P, Eccles J C, Sears T A 1964 Cortically evoked depolarization Neurophysiology 27: 63-77 of primary afferent fibers in the spinal cord. Journal of

Birch L, Christensen H, Arendt-Nielsen L, Graven-Nielsen T, Sf1Igaard K 2000 The influence of experimental muscle pain on motor unit Physiology 83: 200-206 activity during low-level contraction. European Journal of Applied Boyd-Clark L C, Briggs C A, Galea M P 2002 Muscle spindle muscles of the cervical spine. Spine 27: 694-701

Appelberg B, Hulliger M, Johansson H, Sojka P 1983 Actions on muscle afferent fibres in the hind limb of the cat. Journal of Physiology 335: 275-292

gamma-motoneurons elicited by electrical stimulation of group III

distribution, morphology, and denSity in longus colli and multifidus

Arendt-Neilsen L, Graven-Neilsen T, Svarrer H, Svensson P 1996 The during gait: a clinical and experimental study. Pain 64: 231-240 muscle function in fibromyalgia: effect of regional sympathetic 187-191 influence of low back pain on muscle activity and coordination

Brooks D J 1993 Functional imaging in relation to parkinsonian Broton J G, Sessle B J 1988 Reflex excitation of masticatory muscles joint of the cat. Archives of Oral Biology 33: 741-747 syndromes. Journal of Neurological Sciences 115: 1-17

Backman E, Bengtsson A, Bengtsoon M, Lennrnarken C 1988 Skeletal blockade with guanethidine. Acta Neurologica Scandinavica 77: Bernard J F, Bandler R 1998 Parallel circuits for emotional coping behaviour: new pieces in the puzzle. Journal of Comparative Neurology 401: 429-436

induced by algesic chemicals applied to the temporomandibular

Bushnell M C, Duncan G H, Hofbauer R K, Ha B, Chen J-I, Carrier B cortex? Proceedings of the National Academy of Sciences USA 96: Bly N N, Merzenich M M, Jenkins W M 1996 A primate 7705-7709 1999 Pain perception: is there a role for primary somatosensory

genesis model of focal dystonia and repetitive strain inju ry.

1 84


I: Learning-induced dedifferentiation of the representation of the hand in the primary somatosensory cortex in adult monkeys. Neurology 76: 508-520 Bly N N, Merzenich M M, Cheung S, Bedenbaugh P, Nagarajan S S, Jenkins W M 1997 A primate model for studying focal dystonia and repetitive strain injury: effects on primary somatosensory cortex. Physical Therapy 77(3): 269-284 Carpenter D, Lundberg A, Norrsell U 1963 Primary afferent depolarization evoked from the sensorimotor cortex. Acta Physiologica Scandinavica 59: 126-142 Cavada C, Goldman-Rakic P S 1989 Posterior parietal cortex in rhesus monkey. II: Evidence for segregated corticocortical networks linking sensory and limbic areas with the frontal lobe. Journal of Comparative Neurology 287: 422-445 Cheema S S, Rustioni A, Whitsel B L 1984 Light and electron microscopic evidence for a direct corticospinal projection to superficial laminae of the dorsal hom in cats and monkeys. Journal of Comparative Neurology 225: 276-290 Cobb C R, de Vries H A, Urban R T, Luekens C A, Bagg R J 1975 Electrical activity in muscle pain. American Journal of Physical Medicine 54: 80-87 Coghill R C, Talbot J D, Evans A C, et al 1994 Distributed processing of pain and vibration by the human brain. Journal of Neuroscience 14: 4095-4108 Colebatch J G, Deiber M-P, Passingham R E, Friston K J, Frackowiak R S J 1991 Regional cerebral blood flow during voluntary arm and hand movements in human subjects. Journal of Neurophysiology 65: 1 392-1401 Cote L, Crutcher M D 1991 The basal ganglia. In: Kandel E R, Schwartz J H, Jessell T M (eds) Principles of neural science, 3rd edn. Elsevier, New York, pp 647-659 Coulter J D, Jones E G 1977 Differential distribution of corticospinal projections from individual cytoarchitectonic fields in the monkey. Brain Research 129: 335-340 Cowan S M, Bennell K L, Hodges P W, Crossley K M, McConnell J 2001 Delayed onset of electromyographic activity of vastus medialis obliqlllls relative to vastus lateralis in subjects with patellofemoral pain syndrome. Archives of Physical Medicine and Rehabilitation 82: 183-189 Craig A D, Hepplemann B, Schaible H G 1988 The projection of the medial and posterior articular nerves of the cat's knee to the spinal cord. Journal of Comparative Neurology 276: 279-288 Devinsky 0, Morrell M J, Vogt B A 1995 Contributions of anterior cingulate cortex to behaviour. Brain 118: 279-306 Djupsjbbacka M, Johansson H, Bergenheim M 1994 Influences on the y-muscle spindle system from muscle afferents stimulated by increased intramuscular concentrations of arachidonic acid. Brain Research 663: 293-302 Djupsjbbacka M, Johansson H, Bergenheim M, Wenngren B I 1995 Influences on the y-muscle spindle system from muscle afferents stimulated by increased intramuscular concentrations of bradykinin and 5-HT. Neuroscience Research 22: 325-333 Dum R P, Strick P L 1996 Spinal cord terminations of the medial wall motor areas in macaque monkeys. Journal of Neuroscience 16: Durette M R, Rodriguez A A, Agre J C, Silverman J L 1991 Needle electromyographic evaluation of patients with myofascial or fibromyalgic pain. American Journal of Physical Medicine 70: 154-156 Eccles J C, Lundberg A 1959 Synaptic action in motoneurones by afferents which may evoke the flexion reflex. Archives Italiennes de Biologie 97: 1 99-221 Fahrer H, Rentsch H U, Gerber N J, Beyeler C, Hess C W, Grunig B 1988 Knee effusion and reflex inhibition of the quadriceps: a bar to effective retraining. Journal of Bone and Joint Surgery 70: 635-638 Fetz E E 1968 Pyramidal tract effects on interneurons in the cat lumbar dorsal hom. Journal of Neurophysiology 3 1 : 69-80 6513-6525

Fischer-Rasmussen T, Krogsgaard M, Jensen D B, Dyhre-Poulsen P 2001 Inhibition of dynamic thigh muscle contraction by electrical stimulation of the posterior cruciate ligament in humans. Muscle Nerve 24: 1482-1488 Fox J L 1984 The brain's way of keeping touch. Science 225: 820-821 Frank J S, Earl M 1990 Coordination of posture and movement. Physical Therapy 70: 855-863 Galea M P 2002 The neuroanatomy of pain. In: Strong J, Unruh A, Wright A, Baxter G D (eds) Pain: a textbook for therapists. Churchill Livingstone, Edinburgh Galea M P, Darian-Smith I 1994 Multiple corticospinal neuron populations in the macaque monkey are specified by their unique cortical origins, spinal terminations and connections. Cerebral Cortex 4: 1 66-194 Galea M P, Darian-Smith I 1997 Corticospinal projection patterns following unilateral cervical spinal cord section in the newborn and juvenile macaque monkey. Journal of Comparative Neurology 381: 282-306 Gardner E 1950 Reflex muscular responses to stimulation of articular nerves in the cat. American Journal of Physiology 161: 133-141 Gladden M H, Jankowska E, Czarkowska-Bauch J 1998 New observations on coupling between group II muscle afferents and feline y motoneurons. Journal of Physiology 512: 507-520 Gracely R H 1994 Studies of pain in normal man. In: Wall P D, Melzack R (eds) Textbook of pain. Churchill-Livingstone, Edinburgh, pp 315-355 Graven-Nielsen T, Svensson P, Arendt-Nielsen L 1997 Effects of experimental muscle pain on muscle activity and co-ordination during static and dynamiC motor function. Electroencephalography and Clinical Neurophysiology 105: 156-164 Griliner S 1975 Locomotion in vertebrates. Physiological Reviews 55: He X, Proske U, Schaible H-G, Schmidt R F 1988 Acute inflammation of the knee joint in the cat alters responses of flexor motoneurons to leg movements. Journal of NeurophYSiology 59: 326-340 Hides J A, Stokes M J, Saide M, Jull G A, Cooper D H 1994 Evidence of lumbar multifidus muscle wasting ipsilateral to symptoms in patients with acute/subacute low back pain. Spine 19: 165-172 Hodges P W, Richardson C A 1996 Inefficient muscular stabilization of the lumbar spine associated with low back pain. Spine 21 : 2640-2650 Hodges P W, Richardson C A 1 998 Delayed postural contraction of transversus abdominis in low back pain associated with movement of the lower limb. Journal of Spinal Disorders 1 1 : 46-56 Hodges P W, Richardson C A 1999 Altered trunk muscles recruitment in people with low back pain with upper limb movement at different speeds. Archives of Physical Medicine and Rehabilitation 80: 1005-1012 Hoheisel U, Lehmann-Wilienbrock E, Mense 1989 Termination patterns of identified group II and ill afferent fibres from deep tissues in the spinal cord of the cat. Neuroscience 28: 495-507 Hopkins J T, Ingersoll C D, Edwards J E, Cordova M L 2000 Changes in soleus motoneuron pool excitability after artificial knee joint effusion. Archives of Physical Medicine and Rehabilitation 81 : 1199-1203 Horak F 1991 Assumptions underlying motor control for neurologic rehabilitation. In: Lister M (ed) Contemporary management of motor control problems. Proceedings of the II Step Conference. Foundation for Physical Therapy USA, Alexandria, Virginia, pp 11-27 Horak F, Nashner L 1986 Central programming of postural movements: adaptation to altered support surface configurations. Journal of Neurophysiology 55: 1369-1381 Jacobsen S, Danneskiold-Samsoe B 1987 Isometric and isokinetic muscle strength in patients with fibrositis syndrome: new characteristics for a difficult definable category of patient. Scandinavian Journal of Rheumatology 16: 61-65 247-304

The effect of pain on motor control


Johansson H, Sojka P 1991 Pathological mechanisms involved in

genesis and spread of muscular tension in occupational muscle pain and in chronic musculoskeletal pain syndromes: a hypothesis. Medical Hypotheses 35(3): 196-203

Maier M A, Davis J N, Armand J et al 1997 Comparison of cortico­

motoneuronal (CM) connections from macaque motor cortex and 1274

supplementary motor area. Society for Neuroscience Abstracts 23:

Johansson H, Sjolander P, Sojka P, Wadell I 1989 Different fusimotor assessed in the same primary muscle spindle afferents. Journal of Johansson H, Djupsjobacka M, Sjolander P 1993 Influences on the KCI and lactic acid. Neuroscience Research 16: 49-57 Physiology (Paris) 83: 1-12 reflexes from the ipsi- and contralateral hindlimbs of the cat

Martin H A, Arendt-Nielsen L 2000 Effect of muscle pain and

intrathecal AP-5 on electromyographic patterns during treadmill walking in the rat. Progress in Neuro-Psychopharmacology and Biological Psychiatry 24: 1 1 51-1175

gamma-muscle spindle system from muscle afferents stimulated by

Massion J, Hugon M, Wiesendanger M 1982 Anticipatory postural unloading in man. Pflligers Archives 393: 292-296

changes induced by active unloading and comparison with passive

Jovanovic K, Anastasijevic R, Vuco J 1990 Reflex effects on y

diameter muscle afferents in decerebrate cats. Brain Research 521: Kaas J H 1992 The reorganization of sensory and motor maps in adult mammals. In: Gazzaniga M (ed) The cognitive neurosciences. MIT Press, Cambridge, Massachusetts 89-94

fusimotor neurones of chemically induced discharges in small­

Matre D A, Sinkjaer T, Svensson P, Arendt-Nielsen L 1998 75: 331-339

Experimental muscle pain increases the human stretch reflex. Pain

Melzack R, Casey K L 1968 Sensory, motivational and central control Thomas, Springfield, Illinois, pp 423-443 determinants of pain. In: Kenshalo D R (ed) The skin senses. C C

Kankaanpaa M, Taimela S, Laaksonen D, Hanninen 0, Airaksinen 0 patients and controls. Archives of Physical Medicine and Rehabilitation 79(4): 412-417 1998 Back and hip extensor fatigability in chronic low back pain

Mense S 1986 Slowly conducting fibres from deep tissues: neurological properties and central nervous action. Progress in Sensory Physiology 6: 139-219

Mense S 1991 Considerations concerning the neurobiological basis of muscle pain. Canadian Journal of Physiological Pharmacology 69: 610-616

Kellgren J H 1938 Observations on referred pain arising from muscle. Kelso J A S 1995 Dynamic patterns: the self-organization of brain and Kelso J A S, Tuller B 1984 A dynamical basis for action systems. In: Press, New York, pp 321-356 behaviour. MIT Press, Cambridge, Massachusetts Clinical Sciences 3: 175-190

Mense S, Meyer H 1985 Different types of slowly conducting afferent 403-417

units in cat skeletal muscle and tendon. Journal of Physiology 363:

Gazzaniga M (ed) Handbook of cognitive neuroscience. Plenum

Mense S, Skeppar P 1991 Discharge behaviour of feline garnma­ 201-210

motorneurons following induction of an artificial myositis. Pain 46:

Koyama T, Tanaka Y Z, Mikami A 1998 Nociceptive neurons in NeuroReport 9: 2663-2667.

macaque anterior cingulate activate during anticipation of pain.

Merzenich M M, Kaas J H, Wall J T, Nelson R J, Sur M, Felleman D J and 1 in adult monkeys following restricted deafferentation.

1983 Topographic reorganization of somatosensory cortical areas 3b

Lavender S A, Marras W S, Miller R A 1993 The development of Spine 18: 2097-2105

response strategies in preparation for sudden loading to the torso.

Merzenich M M, Nelson R J, Stryker M P, Cynader M S, Shoppmann A, Zook J M 1984 Somatosensory cortical map changes following digit amputation in adult monkeys. Journal of Comparative Neurology Molin C 1972 Vertical isometric forces of the mandible. Acta Nelson D L, Hutton R S 1985 Dynamic and static stretch responses in in Sport and Exercise 17: 445-450 Odontologia Scandinavica 30: 485-499 224: 591-605

Neuroscience 8: 33--55

Lee W 1980 Anticipatory control of postural and task muscles during Ljubisavljevic M, Anastasijevic R 1994 Fusimotor-induced changes in decerebrate cats. Neuroscience 63: 339-348 Experientia 36: 1398 muscle spindle outflow and responsiveness in muscle fatigue in rapid arm flexion. Journal of Motor Behaviour 12: 185--196

Lund S 1980 Postural effects of neck muscle vibration in man. Lund J P, Stohler C S, Widmer C G 1993 The relationship between pain Vaeroy H, Merskey H (eds) Progress in fibromyalgia and myofascial Lundberg A 1979 Integration in a propriospinal motor centre pain. Elsevier, Amsterdam, pp 311-327 and muscle activity in fibromyalgia and similar conditions. In:

muscle spindle receptors in fatigued muscle. Medicine and Science

Newcomer K L, Jacobson T D, Gabriel D A, Larson D R, Brey R H, An low back pain. Archives of Physical Medicine and Rehabilitation 83: Nouwen A, Bush C 1984 The relationship between EMG O'Sullivan P, Twomey L, Allison G, Sinclair J, Miller K 1997 Altered low back pain. Australian Journal of Physiotherapy 43: 91-98 and chronic low back pain. Pain 20: 109-123 81 6-821 K N 2002 Muscle activation patterns in subjects with and without

controlling the forelimb in the cat. In: Asanuma H, Wilson V J 47-64

(eds) Integration in the nervous system. Igaku-Shoin, Tokyo, pp

Lundberg A, Norrsell U, Voorhoeve P 1962 Pyramidal effects on Physiologica Scandinavica 56: 220-229

patterns of abdominal muscle activation in patients with chronic

lumbosacral interneurons activated by somatic afferents. Acta

Pain tal A S 1961 Participation by pressure-pain receptors of (London) 156: 498-514

Lundberg A, Malmgren K, Schomburg E D 1987 Reflex pathways from group IT muscle afferents. 2: Functional characteristics of reflex pathways to alpha-motoneurons. Experimental Brain Research 65:

mammalian spindles in the flexion reflex. Journal of Physiology

Pedersen J, Sjolander P, Wenngren B I, Johansson H 1997 Increased fusimotor drive to muscle spindles in neck muscles of the cat. Pain Price R F, Dutia M B 1989 Physiological properties of tandem muscle 80: 47-57 70: 83--91 intramuscular concentration of bradykinin increases the static

McCrea D A 1994 Can sense be made of spinal interneuron circuits? In: Cordo P, Hamad S (eds) Movement control. Cambridge University Press, New York, pp 31-41


Madeleine P Voigt M, Arendt-Nielsen L 1999a Reorganisation of , and Posture 10: 240-247

spindles in neck and hind-limb muscles. Progress in Brain Research

human step initiation during acute experimental muscle pain. Gait

Radebold A, Cholewicki J, Panjabi M M, Patel T C 2000 Muscle response pattern to sudden trunk loading in healthy 25: 947-954 individuals and in patients with chronic low back pain. Spine

Madeleine P, Lundager B, Voigt M, Arendt-Nielsen L 1999b Shoulder shoulder pain: an occupational pain study. European Journal of

muscle co-ordination during chronic and acute experimental neck­

Applied Physiology and Occupational Physiology 79: 127-140

Radebold A, Cholewicki J, Polzhofer G K, Greene H S 2001 Impaired postural control of the lumbar spine is associated with delayed



muscle response times in patients with chronic idiopathic low back Ralston D D, Ralston H J 1985 The termin.ations of corticospinal tract 242: 325-337 pain. Spine 26(7): 724-730 Stacey M J 1969 Free nerve endings in skeletal muscle of the cat. Stohler C S, Zhang X, Lund J P 1996 The effect of experimental jaw Stohler C S, Kowalski C J, Lund J P 2001 Muscle pain inhibits Sugiura Y, Terui N, Hosoya K 1989 Distribution of unmyelinated cutaneous touch perception. Pain 92: 327-333 muscle pain on postural muscle activity. Pain 66: 215-221 Journal of Anatomy 105: 231-254

axons in the macaque monkey. Journal of Comparative Neurology

Richmond F J, Abrahams V C 1975 Morphology and distribution of Neurophysiology 38: 1322-1339 muscle spindles in dorsal muscles of the cat neck. Journal of

Rossi A, Decchi B, Ginanneschi F 1999 Presynaptic excitability changes Brain Research 818: 1 2-22 of group 1a fibres to muscle nociceptive stimulation in humans.

Headley P M (eds) Processing of sensory information in superficial dorsal hom of the spinal cord. Plenum Press, New York, pp 15-27

primary afferent fibres in the dorsal horn. In: Cervero F, Bennett G J,

Ruch T C 1946 Visceral sensation and referred pain. In: Fulton J F (ed) pp 385-401

Talbot J D, Marrett S, Evans A C, Meyer E, Bushnell M C, Duncan G H Science 251: 1355-1358. 1991 Multiple representations of pain in human cerebral cortex.

Howell's Textbook of physiology, 15th edn. Saunders, Philadelphia,

Rybicki K J, Waldrop T G, Kaufman M P 1985 Increasing gracilis IV afferents. Journal of Applied Physiology 58: 936-941 Psychological Review 94: 84-106

Thorstensson A, Arvidson A 1982 Trunk muscle strength and low back Thunberg J, Ljubisavljevic M, Djupsjobacka M, Johansson H 2002 Effects on the fusimotor-muscle spindle system induced by pain. Scandinavian Journal of Rehabilitation Medicine 14: 69-75

muscle interstitial potassium concentrations stimulate group ill and

Saltzman E, Kelso J A S 1987 Skilled actions: a task-dynamic approach. Schaible H G, Schmidt R F 1985 Effects of an experimental arthritis on Neurophysiology 54: 1109-1122 the sensory properties of fine articular afferent units. Journal of

Travell J G, Simons D G 1983 Myofascial pain and dysfunction: the Treede R-D, Kenshalo D R, Gracely R H, Jones A K P 1999 The cortical Velarde M G, Normand C 1980 Convection. Scientific American 243: Wall P D 1967 The laminar organization of dorsal hom and effects of Wall P D, Woolf C J 1984 Muscle but not cutaneous C-afferent input in the rat. Journal of Physiology 356: 443-458 descending impulses. Journal of Physiology 188: 403-423 78-93 representation of pain. Pain 79: 105-111 trigger point manuaL Williams and Wilkins, Baltimore

Research 143: 319-326

intramuscular injections of hypertonic saline. Experimental Brain

Schaible H-G, Schmidt R F 1988 T ime course of mechanosensitivity arthritis. Journal of Neurophysiology 60: 1 80-195 changes in articular afferents during a developing experimental

Schouenborg J, Sjolund B H 1983 Activity evoked by A- and C-afferent Journal of Neurophysiology 50: 1108-1121 fibres in rat dorsal horn neurons and its relation to a flexion reflex.

Shakespeare D T, Stokes M, Sherman K P, Young A 1985 Reflex with pain. Clinical Physiology 5: 137-144 New York

inhibition of the quadriceps after meniscectomy: lack of association

produces prolonged increases in the excitability of the flexion reflex

Shepherd G M 1994 Neurobiology, 3rd edn. Oxford University Press, Shepherd G 1998 The synaptic organization of the brain, 4th edn. Sherman R A 1985 Relationships between strength of low back muscle contraction and reported intensity of chronic low back pain. American Journal of Physical Medicine 64: 190-200 Oxford University Press, New York

Wang K, Svensson P, Arendt-Nielsen L 2000 Effect of tonic muscle pain Westberg K-G, Clavelou P, Schwartz G, Lund J P 1997 Effects of on short-latency jaw-stretch reflexes in humans. Pain 88: 189-197

chemical stimulation of masseter muscle nociceptors on trigeminal the rabbit. Pain 73: 295-308

motoneuron and interneuron activities during fictive mastication in

Willis W D 1989 Neural mechanisms of pain discrimination. In: Lund J Press, New York, pp 130-143

Sherrington C S 1910 Flexion reflex of the limb, crossed extension 196-214

(ed) Sensory processing in the mammalian brain. Oxford University

reflex, and reflex stepping and standing. Journal of Physiology 47:

Wyke B 1981 The neurology of joints: a review of general principles. Young A, Stokes M, nes J F 1987 Effects of joint pathology on muscle. Yu X-M, Mense S 1990 Somatotopical arrangement of rat spinal dorsal 108: 43-47 Clinical Orthopaedics 219: 21-27 Clinics in Rheumatic Diseases 7: 233-239

Sherrington C S 1947 The integrative action of the nervous system, 2nd Simons D G 1990 Muscular pain syndromes. In: Fricton J R, Awad E York, vol 17, pp 1-41 (eds) Advances in pain research and therapy. Raven Press, New edn. Cambridge University Press, Cambridge

hom cells processing input from deep tissues. Neuroscience Letters

Smeulders M J, Kreulen M, Hage J J, Ritt M J, Mulder T 2002 Motor control impairment of the contralateral wrist in patients with and Rehabilitation 81(3): 1 77-181 unilateral chronic wrist pain. American Journal of Physical Medicine

Zedka M, Prochazka A, Knight B, Gillard D, Gauthier M 1999 pain. Journal of Physiology 520(2): 591-604

Voluntary and reflex control of human back muscles during induced

1 87

Chapter 14

The spine and the effect of ageing
K. P. Singer

187 188 189

A general model of the spine and physiological responses to ageing Disc degeneration Osteophytosis
191 191 192

Patterns of age changes in the spine

Ossification within spinal ligaments

Diffuse idiopathic skeletal hyperostosis nodes
193 194 195

Vertebral end-plate lesions and Schmorl's Zygapophysial and costovertebral joint degeneration Scheuermann's disease Scoliosis Kyphosis
195 195 196 197 197

Degenerative spinal curvature anomalies


Osteoporosis and osteoporotic fracture Spinal trauma Conclusion

Intervertebral disc prolapse

the absence of overt pathology, it is often difficult to dis­ tinguish between the effects of age per se and spinal degen­ erative changes, as both often overlap within the same relative grey scale. There are also different opinions as to whether many individuals achieve old age without some evidence of frank degenerative or pathological changes in spinal joints. Most published reviews on the health of the spine are written from a clinical perspective - pathology, radiology, surgery, biochemistry, clinical anatomy and so on - with an absence of large-scale epidemiological surveys to clarify the issue. The major post mortem survey of the ver­ tebral column conducted over the lifetime of the German pathologist Georg Schmorl (Schmorl 1929, Schmorl & Junghanns 1971) and the large population based radiologi­ cal studies of Dodge et al (1970) and Lawrence (1977) all indicate that spinal degenerative changes have an extremely high prevalence in adult populations. However, although all elements of the spine undergo degenerative changes with age, only some features of this process are associated with spinal pathology and pain (Andersson 1998). It is clear that age and degenerative changes both func­ tion to constrain the system against further injury or dam­ age. In the case of the spine, which serves three prime objectives of mobility, stability and protection of neural ele­ ments, excessive static or dynamic loading stress can induce local or regional degenerative change. Consequently, this chapter draws upon literature which presents both age and degeneration models and their influences upon the spine. There are relatively few reports which document the pat­ terns of age-related degenerative change occurring through­ out a single vertebral column; most focus on either the lumbar or the cervical region. Of clinical interest is the well recognized tendency for stress to accumulate at the transi­ tional zones between regions with resultant symptoms aris­ ing from dysfunction or trauma and their degenerative sequelae. As the literature on normal ageing and disease of the human spine is extensive, this summary draws upon selected reviews and a general model of the vertebral

1 88


column upon which regional patterns of degeneration may be viewed. Reference is made to literature to illustrate degenerative conditions which are represented in specific areas of the spine.

Normal physiological strains are well accommodated by each functional mobile segment; this comprises an interver­ tebral disc (IVD) with an annulus fibrosus and nucleus pul­ posus and the vertebral end-plates (VEP) which include the bony epiphyseal rim at the periphery of the vertebral bod­ ies to regulate their circumferential and vertical growth (O'Rahilly et al 1980). The thin superior and inferior carti­ laginous end-plates connect with the subchondral bony lamella which supports the cancellous trabecular structure within the vertebral body. Paired synovial zygapophysial joints link both vertebrae posteriorly and articulate closely to regulate both load and movement of the segment. Applied moments from muscle actions and axial compres­ sive loads may be coupled with shear, bending (rotations) and torsion about the long axis of the spine, which are in tum moderated by the unique geometry of the segment's zygapophysial and ligamentous anatomy (Fig. 14.1). Inertial strains from dynamic loading, even several times the individual's body weight, may also be tolerated by the spine given its unique capacity to attenuate energy (Adams et aI2002).

Figure 14.1 The motion segment is subjected to applied moments from muscle actions and axial loading, which may be coupled with shear, bending and torsion about the long axis of the spine. These forces are in turn moderated by the segment's zygapophysial joints and disposition of the ligamentous anatomy.

The configuration of the nucleus pulposus comprises amorphous mucopolysaccharides, from the fetal period, which continuously differentiate peripherally with the transverse growth of the annulus fibrosus. Formed through multiple fibrous concentric laminar layers, the annulus lay­ ers are orientated 40-90 degrees relative to each other. This arrangement provides considerable resistance to strains from external force moments and from the internal hydraulic pressures exerted by the nucleus as it deforms in response to load. At the superior and inferior surfaces of the disc, fibres penetrate the VEP and also fuse with the ante­ rior and posterior longitudinal ligaments. Throughout life there is a progressive loss of differentiation between the nucleus and the annulus which on T2-weighted MRI scans is seen as a reduction in the high signal from the central disc region. In the younger individual, the nucleus is more gelatinous but this is gradually replaced by fibrocartilage to become more dehydrated and desiccated in the elderly (Prescher 1998). By the third decade of life, it becomes more difficult to distinguish boundaries between these two prin­ cipal IVD elements. The outermost layers of the annulus comprise mostly type I collagen whereas type II collagen is represented to a higher extent in the innermost layers and within the nucleus. The IVD is designed to attenuate dynamic and static loads through hydraulic mechanisms related to its capacity to bind or express water. The proteo­ glycan gel, principally within the nucleus, maintains fluid content under load in tum sustained by the collagen weave of the surrounding disc matrix which permits deformations in response to physiological motions. With ageing there is a reduction in the nucleus capacity to bind water which is demonstrated through imaging studies as an apparent reduction in vertical disc height, with 'fissures' in the region of the nucleus. This trend has also been demon­ strated mechanically from ex vivo assessments of axial load distributions using disc profilometry, whereby greater loading occurs preferentially through the region of the annulus with a relative decompression in the region of the dehydrated nucleus (Adams et a12002, McNally et aI1992). The reduction in elasticity of the disc can contribute to an increase in the transfer of compressive loads to the YEP, leading to subchondral trabecular microfractures (Hahn et aI1994), a process which may contribute to sclerosis of the end-plate. Radiological demonstration of such changes may be seen in MRI investigations of the spine as the increased signal from T2-weighted image sequences described by Modic et al (1988). The disc is essentially aneural apart from the peripheral superficial outer third, although with injury to the disc, vas­ cular ingrowth associated with repair may contribute vaso­ motor nerves (Coppes et aI1997). The disc is also avascular, apart from the peripheral annulus, with a reliance upon nutritional substances transported via diffusion across the YEPs (Roberts et al 1997) or through vessels which com­ municate directly with the outer annular layers. Consequently, disruption to either system, occurring

The spine and the effect of ageing

1 89

through normal ageing, surgical intervention, spinal defor­ mity or trauma, can disrupt and lengthen the pathways of nutritional support to the disc and is presumed to con­ tribute to subsequent disc degeneration (Buckwalter 1995, Urban & Roberts 1995). In the case of the VEP, sclerosis associated with end-plate lesions or coincident with ageing would likely contribute to this degenerative sequelae of the disc. The consequence of either ageing or injury to the func­ tional mobile segment may be degeneration of its elements with initial progressive increase in strain tolerance beyond the normal, which may progress to increased segmental mobility. One mechanical response to such changes, partic­ ularly affecting the stability and function of the IVD, is spondylosis, initiated through osteogenic stimulation in the junctional region between the VEP periphery and the annulus, resulting in the early formation of osteophytes (Vernon-Roberts & Pirie 1977). Experimentally induced osteoarthrosis of the paired zygapophysial joints, arising from annular rim lesions of the IVD, which has been pro­ duced in the sheep model (Moore et al 1999) confirms ear­ lier post mortem observations (Oegema & Bradford 1991, Osti et al 1992). The posterior paired zygapophysial, costotransverse and costovertebral joints are true synovial joints invested with hyaline articular cartilage, a capsule and synovium. These joint pairings contribute stability of the respective seg­ ment(s), and facilitate respiratory excursions of the thorax and regional mobility within the vertebral column. Each may respond to undue strain with typical degenerative pat­ terns of synovial joints characterized by mechanical changes of the articular cartilage. Subchondral bone sclero­ sis, fissuring and detachment of the cartilage and marginal joint osteophytosis may follow changes in the IVD, particu­ larly a loss of vertical height which in turn alters the mechanical alignment of the respective superior and infe­ rior articular processes of the posterior joints, contributing to subluxation (Oegema & Bradford 1991). Bumper carti­ lage formations are associated with evidence of articular cartilage degeneration and fissuring, ossification of the lig­ amentum flavum and reactive hyperplasia at the posterior joint margins (see Ch. 3, Fig. 3.3). A further consequence of degenerative changes leading to altered morphology of the IVD and vertebral bodies is the response by the spinal liga­ ments. With progressive deformation of the segment, liga­ ments may demonstrate buckling and, in response to exaggerated segmental motion strains, subsequent hyper­ trophic changes may contribute to stenotic change within the vertebral and intervertebral canals (Benini 1990, Weinstein et al 1977). Considerable ossification within the ligamentum flavum may occur as part of degeneration of the articular triad, although this tends to predominate in the region of the lower thoracic and upper lumbar seg­ ments (Maigne et al 1992, Malmivaara 1988). Against this background, it is possible to examine spe­ cific patterns of degeneration and age changes as they are

represented from reported surveys. Indeed, when spinal degenerative patterns are merged onto a common spinal model, the most mobile cervical and lumbar segments and their respective stiffer transitional junctions can be clearly and differentially identified. For the entire spine, there is a tendency for large segmental mobility to induce local strains and ultimately degenerative sequelae. In the case of the bony thorax, high levels of degenerative change are seen at the respective costovertebral joint articulations of the first and last ribs as a consequence of the muscle attach­ ments and the transfer of large torques from this muscula­ ture of the neck and trunk, respectively. W hen one considers the complete vertebral column as a multiseg­ mented curved rod, with physiological inflexions that cross the neutral axis line, the literature presents evidence of stress accumulations at points of both maximum and mini­ mum change in curve. The transitional junctions, having less relative motion, are designed more for stability. They represent locations where axial compressive load is greater, the change in spinal curvature is least and the trend is for arthrosis of their synovial joints. In contrast, where the cur­ vature away from the neutral axis line is maximum, as in the middle region of the lordosis and kyphosis, and where bending, torsion and shear stresses are relatively higher, the trend is for greater disc degeneration (Fig. 14.2). A mixed pattern of degenerative change within a mobile segment emerges where an advanced level of either arthrosis or IVD degeneration has developed.

The major degenerative conditions reviewed in this chapter are listed in Box 14.1 and include osteoporosis and anom­ alies of spinal curvature, and changes which arise second­ ary to trauma. Inflammatory disease of the spine is excluded from this discussion; the interested reader is directed to the compilation by Klippel & Dieppe (1998) for a thorough review. Degenerative conditions which princi­ pally have a spinal manifestation may involve various ele­ ments of the functional mobile segment, either singularly as in the case of early IVD degeneration or across this joint complex, exemplified by late zygapophysial joint arthrosis coincident with IVD degeneration (Fujiwara et aI2000).
Disc degeneration

Literature describing the incidence of disc degeneration (DD) throughout the vertebral column concentrates pre­ dominantly on the lumbar and cervical regions of the spine (Kramer 1981). An active debate has existed on the aetiol­ ogy of DD, from which two main models emerge. Mechanical strains of the annulus fibrosus may result in a rim lesion which initiates this degenerative sequence (Osti et al1990, 1992), or a lesion to the VEP may affect the discs' nutrition. The latter pathology has been confirmed experi­ mentally (Ariga et a12001) and demonstrated clinically and



Box 14.1

Major types of degenerative and ageing

• D� •

effects in the human spine including those related to metabolic disease or arising in response to trauma

D �.

Degenerative conditions Degenerative disc disease, osteophytosis and ossification Diffuse idiopathic skeletal hyperostosis Vertebral end-plate lesions and Schmorl's nodes Scheuerman n 's disease Zygapophysial and costovertebral join t osteoarthritis Spinal alignment anomalies Degenerative scoliosis and kyphosis, spondylolisthesis Osteoporosis and osteoporotic fracture
Trauma Spinal fractures i n cluding the transitional j unctions Intervertebral disc inj ury and

D . � D[!] D D� . � [!]


Figure 14.2 The human vertebral column comprises a balanced 'S'-shaped multisegmented curved rod with physiological inflexions that cross the neutral axis line. Evidence of stress accumulations is seen at points of both maximum and minimum change along this curve. The transitional junctions are constrained against motion, sustain higher axial compressive loads and show high levels of arthrosis of their synovial joints. Where the spinal segments depart from the neutral axis line in the middle lordoses and kyphosis, and where bending, torsion and shear stresses are relatively higher, the tendency is for higher levels of disc degeneration. Key: AOJ = atlanto-axial junction; CTJ = cervicothoracic junction; TLJ thoracolumbar junction; LSJ = lumbosacral junction; OA = osteoarthritis; OP = osteophytes; DD = disc degeneration; EPL end-plate lesions.
= =

at post mortem (Pfirrmann & Resnick 2001). From post mortem studies, discs with altered vascularity during the second decade of life show precursor changes to early degeneration (Boos et aI2002). The pathway of age-related degeneration change has been described by Buckwalter as compromised nutrition, loss of viable cells, cell senescence, post-translational modification of matrix proteins, accumu­ lation of degraded matrix molecules, a reduction in pH lev­ els that may impede cell function and ultimately induce cell death, and finally, fatigue failure of the matrix (Buckwalter 1995). Radiological reviews of large populations under­ taken by Dodge et al (1970) and Lawrence (1977) have indi­ cated that the highest prevalence of DO is in the

mid-cervical, mid-thoracic and mid-lumbar discs. These regions show a marked degree of reactive changes of the vertebral bodies with marginal osteophyte formation (Fig. 14.3). The incidence of DO was commonly linked to occu­ pation and gender; with a greater frequency in males. In the survey by Bobko et al, it was noted that manual labourers were susceptible to DO within the cervicothoracic junction (CTJ) region in contrast to white collar workers where DO was more common within the C4-5 and C5-6 discs (Bobko et aI1966). Early post mortem studies of von Lushka (1850, 1858) demonstrated a large proportion of cervical discs with fissures and clefts. This was considered to be a normal characteristic of the region, an observation which was sub­ sequently confirmed by Tondury (1959) and other workers (Bland 1987, Penning 1988). Complete transverse clefts which extend across and into the region of the uncoverte­ bral joints may be found in the middle of healthy cervical discs on coronal inspection (Ten Have & Eulderink 1980). The high frequency of degenerative findings in the mid­ cervical spine is well documented (Fletcher et al 1990, Milne 1991) and appears to relate to the combination of disc facet and interfacet angles seen in the mid-cervical verte­ brae. Large anteroposterior translation during sagittal motion and combined lateral flexion and axial rotation (Milne 1991, 1993) is thought to result in greater shear forces in the intervertebral discs. The variations in orienta­ tion of the zygapophysial joints through the cervicothoracic transition may in part account for the discal degeneration in the upper thoracic spine. Significant trends in degenerative changes in the cervi­ cothoracic transition with respect to age have been identi­ fied (Boyle et al 1998b). The frequency of osteophytic lipping decreases from the mid-cervical mobile segments with the lowest incidence of marginal vertebral osteophyte formation occurring at the C7-T1 vertebral segment (Nathan 1962). From post mortem reviews of the thoracic

The spine and the effect of ageing

19 1

The pattern of age-related decline in anterior disc height in men typifies the disc ageing process associated with senile kyphosis, as described by Schmorl & Junghanns (1971). Hence in older males without marked spinal osteopenia it is speculated that the cumulative effects of axial loading and torsional stresses result in degeneration of the anterior annulus and osteophytosis (Schmorl & Junghanns 1971). This early observation has been con­ firmed in recent series (Goh et al 1999, Manns et al 1996, Resnick 1985). In older women, however, loading through the anterior aspect of the kyphotic curve is more likely to produce progressive change of the vertebral bodies, caus­ ing the wedge deformity commonly associated with spinal osteoporosis. Mechanically, the middle vertebral segments appear predisposed to greater axial compressive and bend­ ing moments, due to their position within the apex of the thoracic kyphosis (Singer et aI1995).

Figure 14.3 Macro-histology of the T10-11 and L2-3 discs sec­ tioned in the horizontal plane at the mid-height of the disc of two elderly cases. The thoracic example (upper) depicts a central disc prolapse deforming the anterior dural sac. Note that the postero­ lateral disc is prevented from prolapse due to the location of the costovertebral joints in a manner analogous to the uncovertebral joints in the cervical spine. In the lower illustration age-related changes are demonstrated in the form of the large right-sided anterolateral osteophyte and central disintegration of the nucleus. A central fissure is evident through the posterior annulus.

spine, the most severely affected discs were located pre­ dominantly within the middle segments, peaking between T6 and T7, with a greater incidence in males (Singer 2000). Given the higher amplitude of axial plane segmental motion in the mid-thoracic spine, reported from in vivo investigations (Gregersen & Lucas 1967), these degenera­ tive changes may relate to rotation strains imposed upon these segments. Investigation into the effects of torsion on lumbar IVDs has concluded that relatively small rotation strains induced potential injury in the annulus fibrosus (Farfan et al 1970). Similarly a torsion induced strain response from the relatively large axial plane motions pos­ sible in the mid-thoracic segments may be a major factor contributing to the DD seen in these segments (Lawrence 1977, Singer 2000).

A review by Nathan (1962) saw osteophyte formation and its associated IVD degeneration as an attempt to distribute force more uniformly across the VEPs and to achieve stress reduction on the segment. In his study of 346 skeletal spinal columns, Nathan (1962) reported a higher incidence of ver­ tebral body osteophytes, including complete fusion between adjoining vertebrae, in the lower thoracic levels. Where thoracolumbar disc degeneration is present, mar­ ginal osteophyte formation of the vertebral body is fre­ quently seen (Lawrence 1977, Vernon-Roberts 1992). This pattern of excess bone formation is commonly referred to as spondylosis deformans (Resnick 1985) and is seen in approximately 60% of women and 80% of men (Schmorl & Junghanns 1971). In an advanced stage, with complete ossi­ fication of the ligaments from several adjacent vertebrae, this presentation may form part of diffuse idiopathic skele­ tal hyperostosis (DISH) (Belanger & Rowe 2001). The degree of intervertebral space narrowing and subsequent tilting of the vertebral bodies, resulting from disc degener­ ation, often determines the extent and the type of marginal osteophytes (Malmivaara 1987, Nathan, 1994). In summary, the segments that appear susceptible to osteophytes are often the most mobile regions with the higher levels of DD, or where local stress may be accumulated.
Ossification within spinal ligaments

Ossification of the attachments of the ligamentum flavum, which is considered to be a response to stress, has been reported from several surveys of skeletal vertebral columns (Davis 1955, Maigne et al 1992, Nathan 1959). The high fre­ quency of laminar projections of bone localized in the region of the lower thoracic and thoracolumbar junction (TLJ) levels suggests that this is a normal feature of the region, from the third decade of life. Maigne et al (1992) suggested that the size and frequency of these processes



projecting into the ligamentum flavum acted to regulate the segmental response to torsion of the vertebral column, whereas Prescher (1998) considered this to be a reaction to flexion strain. Maigne et al noted that immediately above the TLJ, where the zygapophysial joints were configured to facilitate rotation, the spicules were more developed (Maigne et aI1992) (Fig. 14.4).
Diffuse idiopathic skeletal hyperostosis

Ankylosing hyperostosis of the spine may result in ossifica­ tion of spinal ligaments without evident disc disease. Typically this condition affects older men, being uncom­ mon in men younger than 40 years, and does not usually result in severe disability (Weinfeld et aI1997). According to surveys by Resnick & Niwayama (1976, 1995) radiographic evidence of DISH may be found in 12% of the population. In the majority of these cases, the thoracic spine is involved (Malone et al 1998), particularly on the right side of the

T5-12 vertebral bodies due to the influence of the descend­ ing abdominal aorta (Resnick & Niwayama 1976). There are few published large-scale studies of the prevalence of DISH; however, a recent investigation of 2364 patients iden­ tified 25% of males and 15% of females over the age of 50 years with radiographic features of this disease (Weinfeld et aI1997). The pathology of DISH also involves bone spur formation (enthesophytes) within peri-articular ligaments of other large joints. The condition may be largely asymptomatic; however, when symptoms are present they commonly consist of spinal stiffness, particularly in the morning, and thoracic back pain, with spinal tenderness reported to be present in up to 90% of cases (Utsinger 1985). Loss of spinal mobility and associated diminished thoracic cage motion has been described (Resnick & Niwayama 1976). The principal fea­ tures consist of multiple consecutive flowing osteophytes along the course of the anterior longitudinal ligament, involving at least four adjacent vertebrae, relatively pre-

Figure 14.4 Photomicrographs of 100 �m-thick horizontal histological sections at Tll-12, to high­ light ossification within the ligamentum flavum (arrows). A, B: The ligament is bounded by attach­ ments to the superior articular process and laminae as it helps forms the dorsal wall of the vertebral canal between the paired zygapophysial joints later­ ally and adjacent laminae. In both, there is an expan­ sion of the ligamentum flavum towards the vertebral canal which, in some cases, contributes to central stenosis. From macerated vertebrae at the thora­ columbar junction. Black arrows depict the location of ossicles between T11 and T12 which project into the ligamentum flavum. C, D: Hypertrophic enlarge­ ment of the superior aspect of the sap is evident on the left zygapophysial joint (Z) in D. Adapted from Singer 2000. Key: sap superior articular process; iap = inferior articular process; LF = ligamentum flavum; MP = mammillary process; Z zygapophysial joint.
= =

The spine and the effect




served disc spaces and VEPs (Fig. 14.5), and the absence of sacroiliac or zygapophysial joint sclerosis or ankylosis. The .coexistence of multiple pathologies with DISH was a feature of the post mortem investigation by Vernon­ Roberts et al (1974). In their study, osteoporosis, Schmorl's nodes and lateral projection of disc tissues were found to be associated with DISH. The thickened syndesmophytes acted to bridge the disc space and maintain disc height, in contrast to other forms of spinal degenerative condition.
Vertebral end-plate lesions and Schmorl's nodes

The vertebral end-plate is a membrane of tissue comprising hyaline cartilage and a thin trabecular layer at the dis­ covertebral junction (Grignon et al 2000). Its role is to medi­ ate axial compressive load applied to the IVD and permit transfer of this energy within the subchondral and cancel­ lous bone of the vertebral body. Cyclic physiological axial loading, as occurs with gait, acts as pump mechanism to assist diffusion of nutrients within the vascular vertebral body across the YEP to the disc. Abrupt or fatigue axial loading of the spine may cause localized failure of the YEP resulting in either a frank sharply demarcated vertebral intra-osseous prolapse, often termed a Schmorl's node (SN), or marked irregularity of the end-plate. The repair process for both lesions often results in bony sclerosis which can significantly impair the normal nutrient exchange to the IVD by extending this diffusion pathway. Several authors have suggested that SNs appear most frequently in areas of Y EP weakness, possibly resulting from congenital deficiency of the cartilage end-plate at the site of the remnant notochord. Another theory suggests that scarring of degenerated blood vessels supplying the juvenile disc, whereby the end-plate thickness is reduced,

increases susceptibility to nuclear extrusion (Begg 1954, Resnick & Niwayama 1978, Schmorl & Junghanns 1971). Schmorl's nodes have been reported to occur during the late teens (Fisk et al 1984), with lesions as frequent in the young as in the older individual (Hilton et al 1976, Schmorl & Junghanns 1971). Cadaveric studies of lumbar spines have indicated that SNs develop at an early age and can exhibit advanced degenerative changes (Vernon-Roberts & Pirie 1977). Previous literature investigating the incidence of SNs and Y EP lesions has focused on lower thoracic and upper lumbar segments. The reported incidence of SNs for these investigations have ranged between 38 and 79%, with the variation related to age, sex and racial characteristics of the study sample and geographical issues. Schmorl's nodes are found most commonly in males and are consid­ ered to be related to a genetic disposition, strenuous occupa­ tions (Schmorl & Junghanns 1971) or sports involving dynamic and violent axial loading as might occur with a heavy landing in flexion (Fisk et al 1984). Using cadaver spines containing T9-12 vertebral bodies, Hilton et al found most SNs located within the TlO-ll and Tll-12 segments with no age-related variation (Hilton et al 1976). These authors reported a 60% incidence of SNs in spines of those under the age of 20, with the youngest 13 years of age (Hilton et al 1976). A higher incidence was found by Malmivaara et al who reported SNs in 29 of 37 male cadaveric specimens (TlO-Ll) aged 21-69 years (Malmivaara et al 1987b). In a post mortem radiographic survey, 64 of 90 spines had evidence of SNs (Singer 1997). Both male and female spines displayed SNs throughout the spine, the majority located within the TlO-L2 segments. All authors agree that the inferior end-plate is more susceptible to infraction (Hilton et al 1976, Malmivaara 1987, Yasuma et al 1988) which implies that the Y EP fails under compression.

Figure 14.5 Marked osteophytic development is shown in this lower thoracic sample with dif­ fuse idiopathic skeletal hyperostosis (DISH). From the anterior aspect (Al. extensive bridging osteophytes are seen spanning multiple adja­ cent levels. The internal disc height is preserved and the kyphotic angulation contributed to by vertebral body wedging. Inspection of the median sagittal view from one cleared vertebral body stained with von Kossa (B) shows the extent of the thickened cortical nature of the osteophytes. In contrast a normal vertebral body without marked focal degeneration shows regu­ lar end-plates and a predominance of vertical trabecular bone (e).




Figure 14.6 Intravertebral protrusions, or Schmorl's nodes, are depicted from several views to highlight their location and extent (A). They may project cranially and/or caudally through the vertebral end-plate (arrows). End-plate irregularities are typically in the lower thoracic spine, as represented by the inferior end-plate of T11 (arrow). A depression on the superior end-plate of a 2 mm-thick bone section from T11 is shown at B with slight sclerosing of the end-plate compared with the regular thin inferior end-plate. A central Schmorl's node at T12 (el, in a 100 IJ.m-thick horizontal histological section shows disc material surrounded by sclerotic bony margins. Multiple Schmorl's nodes are shown at the thoracolumbar junction (D), all approximately in the same location and affecting the inferior vertebral end-plate, a char­ acteristic of Scheuermann's disease. Adapted from Singer 2000. Key: c = spinal cord; d disc; ep end-plate; pll = posterior longitudinal ligament; sn Schmorl's nodes.
= = =

Pfirrmann et al made the observation from their post mortem study that SNs were not necessarily associated with evident 00; however, their series of 100 cases were from an elderly post mortem sample (Pfirrmann & Resnick 2001). Figure 14.6 depicts the typical location and presence of multiple SNs in the lower thoracic segments.
Zygapophysial and costovertebral joint degeneration

There appear to be specific sites within the spine where preferential degeneration of the synovial joints occur. The upper and lower segments of the thoracic region show a tendency for zygapophysial and costovertebral joint degen­ eration according to the skeletal surveys of Shore (1935a, 1935b) and Nathan et al (1964). Similar trends for osteo­ phytic remodelling of the zygapophysial joints of the lum­ bosacral junction have been reported (Cihak 1970, Inman & Saunders 1942, Resnick 1985).This may be due to the design of these elements which provide stability and protection in contrast the adjacent mobile segments which show a corre­ spondingly higher frequency of discal disease (Resnick 1985) (see Fig. 14.1). The development of osteophytes and eventual bony fusion of costovertebral and costotransverse joints in aged vertebral columns was also noted by Schmorl

& Junghans (1971) in their extensive survey of spinal pathology (compare Fig. 3.5, in Ch. 3, Zygapophysial joints). Shore believed that maintaining an erect head posture on a changing thoracic kyphosis induced localized stress on the CTJ (Shore 1935a). The increasing thoracic kypho­ sis, which is particularly evident in the aged female popu­ lation (Fon et a11980, Singer et aI1990), and the resulting alteration in cervical spine curvature (Boyle et al 2002) may dispose the CTJ to degenerative changes. The CTJ and TLJ represent transitional areas between mobile and relatively immobile regions of the spine. At the CTJ, Boyle et al found evident IVO and YEP changes, along with osteophytic formation, which were more pronounced in the mobile segments immediately above the transition (Boyle et a11998a, 1998b). This observation was consistent with the patterns of osteophytic formation noted by Nathan (1962). It was considered that the upper thoracic region and thoracic cage acted to impede intersegmental motion and thus safeguard these levels from marked degeneration (Boyle et al 1998a, 1998b). At the TLJ, Malmivaara and co-workers (Malmivaara '1987, Malmivaara et al 1987a) demonstrated that particular pathologies tended to be concentrated at each segment. The T10-11 segment was characterized by disc degenera-

The spine and the effect of ageing


tion, vertebral body osteophytosis and SNs; the Tl1-12 segment tended to show both anterior and posterior degeneration involving zygapophysial and costovertebral joints, while the T12-Ll joint was characterized primarily by posterior joint degeneration. A comparison of zygapophysial joint orientation with degenerative find­ ings suggested that the posterior elements play a signifi­ cant role in resisting torsional loads. Asymmetry in the zygapophysial joint orientation tended to result in degen­ erative changes occurring mostly on the sagittal facing facet (Malmivaara 1987), an observation also made by Farfan et al at the lumbosacral junction (Farfan et aI1972).
Scheuermann's disease


Scheuermann's disease (SD) is a common problem that affects the adolescent spine although late manifestations may also be seen. Major signs are SNs and several wedged vertebral bodies which may contribute to an accentuated thoracic kyphosis. Although the reported incidence varies widely, SD is found in approximately 10% of the popula­ tion, with males and females affected equally (Bradford & McBride 1987). Age of onset is typically before puberty and a key diagnostic feature is the inability of the individual to correct the thoracic deformity. A genetic link is proposed in the aetiology of SD; however, this is yet to be conclusively defined (Aufdermaur 1981). Pathological areas of growth cartilage within VEPs, reported by Ascani & La Rosa, have been implicated in dis­ ordered endochondral ossification (Ascani & Rosa 1994). The extent of vertebral wedging has been related to growth disturbance and mechanical loading, which together pro­ duce the deformity. Shortening of the hamstring and pec­ toral muscles is a common presenting feature. There may be compensatory changes in the lumbar and cervical lordoses. Back extensor muscle function has been found to be reduced in patients with SD compared with controls (Murray et al 1993). Pain in the interscapular region and at the CTJ is a com­ mon presenting symptom in younger patients with SD, while older individuals report more backache, with a dis­ tribution which implicates the TLJ segments (Ascani & Rosa 1994). Symptoms tend to decline with progressive ossification. A careful follow-up study involving 61 patients reported that the natural history of the disease was benign despite the cosmetic and structural disturbances which pro­ duce a characteristic hyperkyphosis (Murray et aI1993). Heithoff et al (1994) described patients with thoracolum­ bar SD who had associated degenerative DD, suggesting that this was a manifestation of an intrinsic defect in the cartilagenous end-plate, resulting in inadequate nutrition and structural weakness, which initiated early disc degen­ eration. Associated co-morbidities may be osteomalacia, Paget's disease, infection and trauma. In the case of trauma, 10% of cases seen at post mortem showed acute rupture of the VEP (Fahey et aI1998).

Idiopathic scoliosis involves a lateral curvature of the spine which is introduced through a disturbance in the longitudi­ nal growth of the spine. It may occur early in the growth of the child and particularly during the early adolescent years (Weinstein 1994). Four main curve patterns have been iden­ tified: thoracic, lumbar, thoracolumbar and double major curves. Each of these curvature patterns has its own char­ acteristics and predictable end-point (Weinstein 1994). Management of this disorder is based on the skeletal matu­ rity of the patient at the time of assessment in relation to projected curve progression and its association with mechanical compromise, disability, back pain and possibly respiratory complications. The key concerns are skeletal immaturity, particularly related to curves of larger magni­ tude. It is well accepted that the severity of the scoliosis can continue to progress through the life span (Ascani et al 1986, Gillespy et al 1985). Disc degeneration is known to develop due to the often extreme compression and ipsilat­ eral tension strains experienced within wedged scoliotic IVDs. In scoliosis, markedly higher levels of the reducible collagen cross-links have been reported on the convex side of the scoliotic disc (Duance et al 1998). This finding appears to parallel changes observed in IVD in degenera­ tive disease to suggest that these reflect increased matrix remodelling. A cascade of degenerative changes occurs in advanced scoliosis due to the attempt to stabilize against the asymmetric mechanical loads induced by this defor­ mity (Fig. 14.7). A less common form, congenital scoliosis, occurs through defects in formation or segmentation of the verte­ bral column. These vertebral anomalies include: hemiverte­ brae, wedge vertebrae and unilateral bar formations (McMaster 1994). The progression of congenital scoliosis depends upon the location of the vertebral anomaly, its type, the extent of growth imbalance it introduces to the spinal column and the age of the individual at the time of diagnosis.

Reference ranges for thoracic kyphosis include a wide spec­ trum across the age span (Fon et al 1980), in part due to a lack of standard assessment system (Singer et al 1990). Alteration in mechanics of the thoracic spine, secondary to an increased kyphosis, can have clinical implications in terms of respiratory function compromise, pain and long­ term deformity. Postural subsidence with ageing is repre­ sented by a loss of vertical height, anterior deformation of the vertebral body and disc (Goh et a11999) and an increase in thoracic kyphosis (Fig. 14.8), which can in turn con­ tribute to secondary changes within the axial skeleton in terms of structure, mechanics and function. A progressive increase in kyphotic angulation becomes most marked in




I -100

I 100


Figure 14.7 An elderly macerated spinal column depicting severe kyphoscoliosis and marked osteophytic fusion across several segments within the region of the thoracolumbar transition, depicted from a posterior (A) and anterior aspect (B). Note the remarkable osseous degen­ eration and remodelling. C: A surface contour image of a marked scoliosis in an elderly woman, showing the typical rib hump appearance and asymmetry of the thoracic cage.

postmenopausal women (Fig. 14.9), whereas in males the kyphos is less susceptible to change, unless influenced by trauma or metabolic disease (Seeman 1995).
Osteoporosis and osteoporotic fracture

Osteoporosis is an endocrine disease characterized by decreased bone mass and micro-architectural deterioration of bone, which may lead to bone fragility and subsequently to an increased rate of fracture. Although resorption of bone follows the normal process of ageing, it may be induced through disordered metabolism of bone and is accelerated following menopause in women (Kanis 1996). The concern by health economists regarding the full impact of this disease in Western societies is mounting, given the rapidly increasing proportion of older individuals who will require various forms of management of this 'silent' epi­ demic. More recently, osteoporosis in men has become rec­ ognized as having a potentially significant bearing on healthcare costs although the impact may be delayed in onset compared with women (Seeman 2001b). Osteoporotic fractures are uncommon in young adults of either sex as their bone mass, as assessed by dual energy X-ray absorptiometry (DXA), is within normal ranges and bone loading stress is well within acceptable limits. With ageing, a gender difference in bone fragility emerges due to the dynamic change in relationship between the mechanics of load transfer and the margins of safety. Men accumulate more periosteal bone than women with a corresponding

increase in vertebral cross-sectional area which confers a relatively higher loadbearing capability such that reduc­ tions in bone strength are less dramatic than those seen in women. During ageing, this ratio is disturbed and fracture risk increases as the stress on bone begins to approximate its strength. About 20% of postmenopausal women have a stress to strength ratio imbalance, whereas only 2-3% of men are at risk of fracture due to the greater preservation of strength (Seeman 200la). The epidemiology of osteoporosis is well known whereby the risk factors of age, sex and racial contributors to bone loss and corresponding fracture risk increase expo­ nentially with age (Matkovic et al 1995). For the thoracic spine, one in four women over the age of 60 years will show at least one vertebral body fracture on radiographic exami­ nation, while the incidence increases to 100% in women over 80 years of age (Melton 1995). The mid-thoracic segments are the most vulr1erable to osteoporotic collapse or progressive wedge deformity due to the mechanical disadvantage of these segments situated within the apex of the thoracic kyphosis (Edmondston et al 1997). The second peak for thoracic osteoporotic fracture is at the thoracolumbar junction (De Smet et al 1988), where more rapid loading of the thoracic spine can induce a pivot of the stiffened thorax on the TLJ. These more dynamic loads may be sufficient to cause marked collapse fractures under compression. Degenerative change to the IVD is not common in osteoporosis, suggesting a sparing of the IVD despite the vertebral body collapse (Fig. 14.10).

T h e sp i ne and the effect of ageing

Figure 14.9 Schematic illustrati on of the sequential adjustment in cervicothoracic spinal posture over the lifespan. The accentu­ ated thoracic kyphosis tends to reduce the cervical lordosis as a compensation to preserve forward gaze. Reproduced from Boyle et al 2002. Figure 14.8 Radiographic and macroscopic illustrations of the thoracic kyphotic deformity. Marginal osteophytosis is evident at several levels with fusion across the anterior discovertebral junc­ tions (arrow). Slight anterior wedging of multiple vertebral bodies is noted, particularly in the middle segments.

Spinal trauma

Trauma to the spine may act to initiate degenerative changes through inadequate attempt at remodelling and subsequent strains applied to the injured site, or where the effects of ageing further compound the response to injury. The most common mechanisms of spinal fracture of any region involve dynamic flexion and axial loads (Meyer 1992), commonly mediated through falls or motor vehicle injuries (Daffner 1990). Transitional regions of the human spinal column are considered to be particularly vulnerable to injury due to the abrupt changes in vertebral morphol­ ogy which alter spinal mechanics and load transmission (Kazarian 1981, Singer et al 1989). In a study of 2461 spinal fractures, 54% occurred between Tll and L2 (Rehn 1968). The most commonly reported injury at the thoracolumbar junction is the wedge compression fracture (Harkonen et al 1979, Willen et al 1990). While the cervicothoracic junction is not affected by trauma to the same extent, the injuries can be severe when forces are localized to these segments (Evans 1991).

Rapid loading in flexion can induce traumatic Y EP rup­ tures. In a recent post mortem study of 70 spinal trauma cases most acute SNs were identified in spines from indi­ viduals aged between 11 and 30 years (Fahey 1998). The male to female ratio was 9:1, and SNs were predominantly confined to the T8-Ll segments. Of clinical interest was the absence of radiological detection of these acute SN injuries. The coexistence of other spine pathologies, particularly DISH and ankylosing spondylitis, can complicate the man­ agement of patients with spinal fracture (Bernini et al 1981). Traumatic spinal injury, particularly discal lesions, can ini­ tiate subsequent degenerative change through rim lesions (Osti et al 1990) to more subtle disruptions to nutritional pathways via Y EP damage.
I ntervertebral disc prolapse

The entity of disc prolapse was first described in a classic report of paraplegia (Key 1838). Clinically, the regions sus­ ceptible to this injury and the resulting disc degeneration typically are those with higher levels of mobility within the cervical and lumbar regions (Kramer 1981, Kramer et al 1991). What is not often appreciated is the high frequency of macroscopic discal prolapse within the thoracic region (Andrae 1929, Singer 2000) (see Fig. 14.3). Many of the pro-



0 EEl @J �--+--AD

Figure 14.10 Schematic model to illustrate the effect of normal motion segment alignment between pai red vertebral bodies, their paired zygapophysial joints posteriorly and a normal intervertebral disc (A). With advancing years and effect of osteoporosis, the vertebral bodies may demonstrate a reduced vertical height (h). most evident anteriorly, loss of disc height and subluxation of the zygapophysial joints and a slight anterior tilt of the superior vertebrae on the subjacent due to anterior disc wedging (B). An expansion in depth (d) of the vertebral body may be accounted for by endosteal bone deposition and osteophytic formation. The cumulative effect is accentuation of regional spinal curvature and focal deformity where isolated vertebral bodies are wedged through incident fracture.

lapses tend to be small and flat in appearance, an observa­ tion confirmed by Schmorl & Junghanns in their large post mortem series (Schmorl & Junghanns 1971). From a review of seven clinical studies involving 221 cases of surgically confirmed thoracic prolapse, the level within the thoracic spine that showed most frequent protru­ sion was the Tl1-12 (see Fig. 14.3). This may be attributed to the relatively greater disc height and volume at this region, coupled with localization of torsional forces which can occur immediately above the level of TLJ transition. Indeed, Markolf (1972) proposed that the eleventh and twelfth tho­ racic vertebrae represented a site of structural weakness for stresses in the vertebral column, due to the reduced con­ straint of the ribcage and the change in zygapophysial joint morphology which facilitated rotation above the transi­ tional levels and impeded it below. The implication of disc prolapse is decompression of the nucleus, fissuring of the annulus and the cascade of changes which follow this injury (Buckwalter 1995, Singer & Fazey 2004).

practice. This chapter has reviewed the effect of age on the human spine including those degenerative processes which are secondary to metabolic disease, spinal deformity or trauma. Given the projections for an aged population and the prevalence of spinal degenerative conditions, it must be emphasized that patients presenting with any ankylosis and advanced degenerative condition of the spinal column are vulnerable to stress concentration at points of force application. A careful history and appropri­ ate imaging, where indicated, will complement the assess­ ment and assist in determining the appropriateness of manual therapy. Ageing of the spine is not merely a chronological process, as remodelling and repair follow such insults as disease, deformity, trauma or surgery. W hile the spines of some older individuals exhibit rela­ tively few degenerative changes, those that do reflect a fundamental biological strategy to stabilize the segment(s) against further damage from imposed loads (Kirkaldy­ Willis & Farfan 1982). When the spine is seen as a single system with regions of either high stress localization at the transitional junctions or torsion-related strains in regions of mobility, the typical patterns of degenerative responses can be recognized. Age-related physiological responses to abnormal development, trauma and disease can then be superimposed upon this mechanical framework to extend the explanation of how some insults are attenuated better than others. Efforts to disclose further the natural history of disc cell function, to manipulate disc cells and improve disc nutri­ tion, study vertebral end-plate structure and function will assist management of discal disease. Beyond these issues, refinements in tissue engineering, gene therapy and the potential of stem cells in disc therapy will contribute to moderating some of the ageing changes of this critical spinal element (Gruber & Hanley 2003).


The human spine contributes a large proportion of the musculoskeletal presentations seen in manual therapy

spine ageing pathology degeneration intervertebral disc vertebral end-plate zygapophysial joint osteoarth ritis

osteoporosis fracture trauma Schmorl's nodes Scheuermann's disease scoliosis kyphosis disc prolapse

The spine and the effect of agei ng

1 99

Adam� M, Bogduk N, Burton A K, Dolan P 2002 Biomechanics of back pain. Churchill Livingstone, Edinburgh Andersson G B 1998 What are the age-related changes in the spine? Baillieres Clinical Rheumatology 12: 161-173 Andrae R 1929 Uber Knorpelknotchen am hinteren Ende der Wirbelbandscheiben im Bereich des Spinalkanals. Beitrage zur Pathologischen Anatomie und zur Allemeinen Pathologischen 82: 464-474 Ariga K, Miyamoto S, Nakase T et al 2001 The relationship between apoptosis of endplate chondrocytes and ageing and degeneration of the intervertebral disc. Spine 26: 2414-2420 Ascani E, Rosa G L 1994 Scheuermann's kyphosis. In: Weinstein S L (ed) The pediatric spine: principles and practice. Raven Press, New York, pp 557-584 Ascani E, Bartolozzi P, Logroscino C A et al 1986 The natural history of untreated idiopathic scoliosis after skeletal maturity. Spine 1 1 : 784-789 Aufdermaur M, 1981 Juvenile kyphosis (Scheuermann's disease): radiography, histology, and pathogenesis. Clincial Orthopaedics and Related Research 154: 167-174 Begg A C 1954 Nuclear herniations of the intervertebral disc. Journal of Bone and Joint Surgery 36B: 180-193 Belanger T A, Rowe 0 E 2001 Diffuse idiopathic skeletal hyperostosis: musculoskeletal manifestations. Journal of the American Academy of Orthopaedic Surgery 9: 258-267 Benini A 1990 Segmental instability and lumbar spinal canal stenosis. Neurochirurgia 33: 146-157 Bernini P M, Floman Y, Marvel J P, Rothman R H 1981 Multiple thoracic spine fractures complicating ankylosing hyperostosis of the spine. Journal of Trauma 21: 811-814 Bland J 1987 Disorders of the cervical spine. Saunders, Philadelphia Bobko G, Rabai K, Zsadon B 1966 Contribution to the clinical picture of cervical spondylosis. Zeitschrift fur Rheumaforsch 25: 181-186 Boos N, Weissbach S, Rohrbach H, Weiler C, Spratt K F, Nerlich A G 2002 2002 Volvo Award in basic science. Classification of age-related changes in lumbar intervertebral discs. Spine 27: 2631-2644 Boyle J J W, Singer K P, Milne N 1998a Pathoanatomy of the intervertebral discs at the cervicothoracic junction. Manual Therapy 3: 72-77 Boyle J W W, Milne N, Singer K P 1998b Clinical anatomy of the cervicothoracic junction. In: Giles L, Singer K P (eds) Clinical anatomy and management of cervical spine pain. Butterworth Heinemann, Oxford, pp 40-52 Boyle J W W, Milne N, Singer K P, 2002 Influence of age on cervicothoracic spinal curvature: postural implications. Clinical Biomechanics 17: 361-367 Bradford 0, McBride G 1987 Surgical management of thoracolumbar spine fractures with incomplete neurologic deficts. Clinical Orthopaedics and Related Research 218: 201-216 Buckwalter J A 1995 Aging and degeneration of the human intervertebral disc. Spine 20: 1307-1314 Cihak R 1970 Variations of lumbosacral joints and their morphogenesis. Acta Universitatis Carolinae Medica 16: 1 45-165 Cop pes M H, Marani E, Thomeer R T, Groen G J 1997 Innervation of 'painful' lumbar discs. Spine 22: 2342-2349 Daffner R H 1990 Thoracic and lumbar vertebral trauma. Orthopaedic Clinics of North America 21 : 463-482 Davis P 1955 Observations on the movements of the human lower thoracic and lumbar vertebrae. Journal of Anatomy 89: 565 De Smet A A, Robinson R G, Johnson B E, Lukert B P 1988 Spinal compression fractures in osteoporotic women: patterns and relationship to hyperkyphosis. Radiology 166: 497-500 Dodge H J, Mikkelsen W M, Duff I F 1970 Age-sex specific prevalence of radiographic abnormalities of the joints of the hands, wrists and cervical spine of adult residents of the TeclUnseh, Michigan, Community Health Study area, 1962-1965. Journal of Chronic Diseases 23: 151-159 Duance V C, Crean J K, Sims T J et al 1998 Changes in collagen cross­ linking in degenerative disc disease and scoliosis. Spine 23: 2545-2451 Edmondston S J, Singer K, Day R, Price R E, Breidahl W H 1997 Ex­ vivo estimation of thoracolumbar vertebral body compressive strength: the relative contributions of bone denSitometry and vertebral morphometry. Osteoporosis International 7: 142-148 Evans 0 K 1991 Dislocations at the cervicothoracic junction. Journal of Bone and Joint Surgery 65B: 124-127 Fahey V, Opeskin K, Silberstein M, Anderson R, Briggs C 1998 The pathogenesis of Schrnorl's nodes in relation to acute trauma: an autopsy study. Spine 23: 2272-2275 Farfan H F, Cossette J W, Robertson G H, Wells R V, Kraus H 1970 The effects of torsion on the lumbar intervertebral joints: the role of torsion in the production of disc degeneration. Journal of Bone and Joint Surgery 52A: 468-497 Farfan H, Huberdeau R, Dubow H 1972 Lumbar intervertebral disc degeneration: the influence of geometrical features on the pattern of disc degeneration - a post mortem study. Journal of Bone and Joint Surgery 54B: 492-510 Fisk J W, Baigent M L, Hill P 0 1984 Scheuermann's disease: clinical and radiological survey of 17 and 18 year olds. American Journal of Physical Medicine 63: 18-30 Fletcher G, Haughton V, Ho K, Yu S 1990 Age-related changes in the cervical facet joints: studies with cryomicrotomy, MR and CT. American Journal of Roentgenology 154: 81 7-820 Fon G, Pitt M, Thies A 1980 Thoracic kyphosis: range in normal subjects. American Journal of Roentgenology 134: 979-983 Fujiwara A, Lim T H, An H S et al 2000 The effect of disc degeneration and facet joint osteoarthritis on the segmental flexibility of the lumbar spine. Spine 25: 3036-3044 Gillespy T, Gillespy T, Revak C 1985 Progressive senile scoliosis: seven cases of increasing spinal curves in elderly patients. Skeletal Radiology 13: 280-286 Goh S, Price R I, Leedrnan P J, Singer K P 1999 The relative influence of thoracic vertebral and disc morphometry on thoracic kyphosis. Clinical Biomechanics 14: 439-448 Gregersen G, Lucas 0 1967 An in vivo study of the axial rotation of the human thoraco-lumbar spine. Journal of Bone and Joint Surgery 49A: 247-262 Grignon B, Grignon Y, Mainard 0, Braun M, Netter P, Roland J 2000 The structure of the cartilaginous end-plates in elder people. Surgical and Radiologic Anatomy 22: 13-19 Gruber H E, Hanley E N J 2003 Recent advances in disc cell biology. Spine 28: 186-193 Hahn M, Vogel M, Arnling M et al 1994 Mikrokallusformationen der spongiosa. Pathologe 15: 297-302 Harkonen M, Kataja M, Keski-Nisula T et al 1979 Injuries of the thoracolumbar junction. Archives of Orthopaedic and Traumatic Surgery 94: 35-41 Heithoff K B, Gundry C R, Burton C V, Winter R B 1994 Juvenile discogenic disease. Spine 19: 335-340 Hilton R, Ball J, Benn R 1976 Vertebral end-plate lesions (Schrnorl's nodes) in the dorsolumbar spine. Annals of the Rheumatic Diseases 35: 1 27-132 Inman V, Saunders J 0 1942 The clinico-anatomical aspects of the lumbosacral region. Radiology 38: 669-678 Kanis J A 1996 Textbook of osteoporosis. Blackwell Science, Oxford Kazarian L 1981 Injuries to the human spinal column : biomechanics and injury classification. Exercise and Sport Sciences Reviews 9: 297-352 Key C 1838 On paraplegic depending on disease of the ligaments of the spine. Guy's Hospital Report 3: 1 7-34



Kirkaldy-WiLiis W H, Farfan H F 1982 Instability of the lwnbar spine. Clinical Orthopaedics and Related Research 165: 1 1 0-123 Klippel J H, Dieppe P A 1 998 Rhewnatology. Mosby, London Kramer J 1981 Intervertebral disk diseases: causes, diagnosis, treatment and prophylaxis. Georg meme Verlag, Stuttgart Kramer J, Rivera C A, Kleefield J 1991 Degenerative disorders of the cervical spine. Rheumatic Disease Clinics of North America 17: 741-755 Lawrence J S 1977 Rheumatism in populations. Heinemann, London McMaster M J 1994 Congenital scoliosis. In: Weinstein S L (ed) The pediatric spine: principles and practice. Raven, New York, pp 227-244 McNally D S, Adams M A, Goodship A E 1992 Measurement of stress distribution within intact loaded intervertebral discs. In: Little E G (ed) Technology transfer between high tech engineering and biomechanics. Elsevier Amsterdam, pp 139-159 Maigne J Y, Ayral X, Guerin-Surville H 1992 Frequency and size of ossifications in the caudal attachements of the ligamentum flavwn of the thoracic spine: role of rotatory strains in their development. Surgical and Radiologic Anatomy 14: 119-124 Malmivaara A 1987 Disc degeneration in the thoracolwnbar junctional region: evaluation by radiography and discography in autopsy. Acta Radiologica 28: 755-760 Malmivaara A 1988 Thoracolumbar junctional region of the spine: an anatomical, pathological and radiological study [ Abstract]. Acta Radiologica 29: 621 Malmivaara A, Videman T, Kuosma E, Troup J D G 1987a Facet joint orientation, facet and costovertebral joint osteoarthrosis, disc degeneration, vertebral body osteophytosis and Schmorl's nodes in the thoracolumbar junctional region of cadaveric spines. Spine 12: 458-463 Malmivaara A, Videman T, Kuosma E, Troup J D G 1987b Plain radiographic, discographic, and direct observations of Schmorl's nodes in the thoracolumbar junctional region of the cadaveric spine. Spine 12: 453-457 Malone D G, Caruso J R, Baldwin N G 1 998 Degenerative and non­ infectious inflammatory diseases. In: Benzel E C, Stillerman C B (eds) Thoracic spine. Quality Medical Publishing, St Louis, pp 597-606 Manns R A, Haddaway M J, McCall I W, Cassar Pullicino V, Davie M W J 1996 The relative contribution of disc and vertebral morphometry to the angle of kyphosis in asymptomatic subjects. Clinical Radiology 51: 258-262 Markolf K L 1972 Deformation of the thoracolumbar intervertebral joints in response to external loads. Journal of Bone and Joint Surgery 54A: 511-533 Matkovic V, Klisovic D, Ilich J Z 1995 Epidemiology of fractures during growth and aging. Physical Medicine and Rehabilitation Clinics of North America 6: 415-439 Melton J L 1995 Epidemiology of fractures. In: Riggs L B, Melton J L (eds) Osteoporosis: etiology, diagnosis, and management, 2nd edn. Lippincott-Raven, Philadelphia, pp 225-247 Meyer S 1 992 Thoracic spine trauma. Seminars in Roentgenology 27: 254-261 Milne N 1991 The role of zygapophysial joint orientation and uncinate processes in controlling motion in the cervical spine. Journal of Anatomy 178: 1 89-201 Milne N 1993 Composite motion in cervical disc segments. Clinical Biomechanics 8: 193-202 Modic M, Masaryk T, Ross J, Carter J 1988 Imaging of degenerative disk disease. Radiology 168: 177-186 Moore R J, Crotti T N, Osti 0 L, Fraser R D, Vernon-Roberts B 1999 Osteoarthrosis of the facet joints resulting from anular rim lesions in sheep lumbar discs. Spine 24: 519-525 Murray P M, Weinstein S L, Spratt K F 1993 The natural history and long-term follow-up of Scheuermann's kyphOSiS. Journal of Bone and Joint Surgery 75A: 236--248

Nathan H 1959 The para-articular processes of the thoracic vertebrae. Anatomical Record 133: 605-618 Nathan H 1962 Osteophytes of the vertebral column: an anatomical study of their development according to age, race and sex with considerations as to their etiology and significance. Journal of Bone and Joint Surgery 44A: 243-268 Nathan H, Weinberg H, Robin G C, Aviad I 1964 The costovertebral joints: anatomico-clinical observations in arthritis. Arthritis and Rheumatology 7: 228-240 O'Rah ilJy R, Muller F, Meyer D B 1980 The human vertebral column at the end of the embryonic period proper. 1: The column as a whole. Journal of Anatomy 1 3 1 : 565-575 Oegema T R, Bradford D S 1991 The inter-relationship of facet joint osteoarthritis and degenerative disc disease. British Journal of Rheumatology 30: 1 6--20 Osti 0 L, Vernon-Roberts B, Fraser R D 1990 Annulus tears and intervertebral disc degeneration: an experimental study using an animal model. Spine 15: 762-767 Osti 0 L, Vernon-Roberts B, Moore R, Fraser R D 1992 Annular tears and disc degeneration in the lumbar spine: a post-mortem study of 135 discs. Journal of Bone and Joint Surgery 74B: 678-682 Penning L 1988 Differences in anatomy, motion, development and aging of the upper and lower cervical disk segments. Clinical Biomechanics 3: 37-47 Pfirrmann C W A, Resnick D 2001 Schrnorl nodes of the thoracic and lwnbar spine: radiologic-pathologic study of prevalence, characterization, and correlation with degenerative changes of 1650 spinal levels in 100 cadavers. Radiology 219: 368-374 Prescher A 1998 Anatomy and pathology of the aging spine. European Journal of Radiology 27: 181-195 Rehn J 1968 Die knbcheren Verletzungen der Wirbelsaule, Bedeutung des Erstbefundes fiir die spatere Begutachtung. Die Wirbelsaule in Forschung und Praxis 40: 131-138 Resnick D 1985 Degenerative diseases of the vertebral column. Radiology 156: 3-14 Resnick D, Niwayama G 1976 Radiographic and pathologic features of involvement in diffuse idiopathic skeletal hyperostosis (DISH). Radiology 119: 559-563 Resnick D, Niwayama G 1978 Intravertebral disk herniations: cartilaginous (Schrnorl's) nodes. Radiology 126: 57-65 Resnick D, Niwayama G 1995 Diffuse idiopathic skeletal hyperostosiS (DISH): ankylosing hyperostosiS of Forestier and Rotes-Queroi. In: Resnick D (eds) DiagnOSiS of bone and joint disorders. Saunders, Philadelphia, pp 1463-1508 Roberts S, McCall I W, Menage J, Haddaway M J, Eisenstein S M 1997 Does the thickness of the vertebral subchondral bone reflect the composition of the intervertebral disc? European Spine Journal 6: 385-389 Schmorl G 1929 Zur pathologische Anatomie der Wirbelsaule. Klinik Wochenscrift 8: 1243-1249 Schmorl G, Junghanns H 1971 The human spine in health and disease. Grune and Stratton, New York Seeman E 1995 The dilemma of osteoporosis in men. American Journal of Medicine 98(Suppl. 2A): 76--88S Seeman E 2001a During aging, men lose less bone than women because they gain more periosteal bone, not because they resorb less endosteal bone. Calcified Tissue lJlternational 69: 205-208 Seeman E 2001b Unresolved issues in osteoporosis in men. Reviews of Endocrinology and Metabolic Disorders 2: 45-64 Shore L 1935a On osteo-arthritis in the dorsal intervertebral joints: a study in morbid anatomy. British Journal of Surgery 22: 833-849 Shore L R 1935b Some examples of disease of the vertebral column found in skeletons of ancient Egypt: a contribution to paleopathology. British Journal of Surgery 22: 256--271 Singer K P 1997 Pathomechanics of the aging thoracic spine. In: Lawrence D (eds) Advances in chiropractic. Mosby, St Louis, pp 1 29-153

The spine and the effect of ageing


Singer K P 2000 Pathology of the thoracic spine. In: Giles L, Singer K P (eds) Clinical anatomy and management of thoracic spine pain. But!erworth Heinemann, Oxford, pp 63-82 Singer K P, Fazey P J 2004 Disc herniation - non-operative treatment. In: Herkowitz H K et al (eds) The lumbar spine, 3 edn. Raven­ Singer K P, Willen J, Breidahl P D, Day R E 1989 The influence of zygapophyseal joint orientation on spinal injuries at the thoracolumbar junction. Surgical and Radiologic Anatomy 1 1 : 233-239 Singer K P, Jones T J, Breidahl P D 1990 A comparison of radiographic and computer-assisted measurements of thoracic and thoracolumbar sagittal curvature. Skeletal Radiology 19: 21-26 Singer K P, Edmondston S J, Day R E, Price E, Breidahl P B 1995 Prediction of compressive strength of isolated lumbar and thoracic vertebral bodies: relationships to age, geometry, bone density and spinal curvature. Bone 17: 167-174 Ten Have H A M J, Eulderink F 1980 Degenerative changes in the cervical spine and their relationship to its mobility. Journal of Pathology 132: 133-159 Tbndury G 1959 La colonne cervicale, son developpement et ses modifications durant la vie. Acta Orthopedica Belgica 25: 602-627 Urban J P, Roberts S 1995 Development and degeneration of the intervertebral discs. Molecular Medicine Today 1 : 329-335 Utsinger P 1985 Diffuse idiopathic skeletal hyperostosis. Clinics in Rheumatic Diseases 1 1 : 325-351 Vernon-Roberts B 1992 Age-related and degenerative pathology of the intervertebral discs and apophyseal joints. In: Jayson M I V (ed) The lumbar spine and back pain. Churchill Livingstone, Edinburgh, pp 17-42 Lippincott, Philadelphia, in press

Vernon-Roberts B , Pirie C J 1977 Degenerative changes in the intervertebral discs of the lumbar spine and their sequelae. Rheumatology and Rehabilitation 16: 13-21 Vernon-Roberts B, Pirie C J, Trenwith V 1974 Pathology of the dorsal spine in ankylosing hyperostosiS. Annals of the Rheumatic Diseases 33: 281-288 von Luschka H 1850 Die nerven des menschlichen wirbelkanales. Laupp and Siebeck, Tubingen von Luschka H 1 858 Die Halbbgelenke des Menschlichen Kbrpers. Druck und Verlag von Georg Reimer, Berlin, p 144 Weinfeld R M, Olson P N, Maki D D, Griffiths H J 1997 The prevalence of diffuse idiopathic skeletal hyperostosis (DISH) in two large American Midwest metropolitan hospital populations. Skeletal Radiology 26: 222-225 Weinstein S L 1994 Adolescent idiopathic scoliosis: prevalence and natural history. In: Weinstein S L (ed) The pediatric spine. Raven, New York, pp 463-478 Weinstein P, Ehni G, Wilson C 1977 Clinical features of lumbar spondylosis and stenosis. In: Weinstein P R, Ehni G, Wilson C B (eds) Lumbar spondylosis, diagnosiS, management and surgical treatment. Year Book Medical Publisher, Chicago Willen J, Anderson J, Tomooka K, Singer K 1990 The natural history of burst fractures in the thoracolumbar spine (T12 & Ll). Journal of Spinal Disorders 3: 39-46 Yasuma T, Saito S, Kihara K 1988 Schmorl's nodes: correlation of X-ray and histological findings in postmortem specimens. Acta PatholOgical Japonica 38(6): 723-733




Clinical sciences for manual therapy of the spine

SECTION CONTENTS 15. How inflammation and minor nerve injury contribute to pain in nerve root and peripheral neuropathies 16. Chronic pain and motor control 17. Cervical vertigo 233 243 257 215 205

18. The cervical spine and proprioception

19. The vertebral artery and vertebrobasilar insufficiency

20. Mechanisms underlying pain and dysfunction in whiplash associated disorders: implications for physiotherapy


275 291 311

21. The cervical spine and headache

22. 'Clinical instability' of the lumbar spine: its pathological basis, diagnosis and conservative management 23. Abdominal pain of musculoskeletal origin 24. Osteoporosis 347 333





How inflammation and minor nerve injury contribute to pain in nerve root and peripheral neuropathies
J. Greening

205 205 206 207

Minor peripheral nerve injury Peripheral neuritis neuropathic pain
207 208

The role of cytokines and neurotrophins in


Inflammation and compression of nerve roots and dorsa I root 9an9 lion
208 209

Disc herniation: effects on nerve roots to the dorsa I root 9an9 I ion Conclusion
211 210

Comparisons between injury proximal or distal Nerve roots and trauma: whiplash injuries

Spinal and peripheral nerve injury involving axonal degen­ eration is known to cause neuropathic symptoms associ­ ated with sensory and motor fibre changes. These conditions are easy to diagnose when associated with objec­ tive loss of nerve conduction velocity (NCV) or muscle weakness. However, many patients present without objec­ tive signs of nerve dysfunction but with symptoms sugges­ tive of neuropathy (paraesthesia, spontaneous pain, lancinating, burning pain). Examples are non-specific back pain, chronic whiplash disorders and non-specific arm pain. Patients may complain of pain that is not in a der­ matomal, myotomal or single peripheral nerve distribution and symptoms may be associated with hyperalgesic or allo­ dynic sensory changes. Over the past decade neurophysiol­ ogists and clinicians have made considerable progress towards understanding how minor nerve injury or inflam­ mation without major axonal degeneration can cause changes similar to frank nerve injury with altered sensory thresholds and neuropathic symptoms. This chapter will review the evidence for altered nerve function, morphology and neuropathic symptoms follow­ ing minor peripheral nerve and nerve root injury. The effects of inflammatory mediators and mechanical com­ pression on these structures will be examined and related to clinical studies where possible. While most of this work has been carried out in animals using large peripheral nerves, usually the sciatic, some of the neuropathological mecha­ nisms involved are similar to experimental models of nerve root injury. Clinical studies confirm that these animal mod­ els of nerve injury may be relevant to the many patients musculoskeletal physiotherapists see with these chronic pain problems. Table 15.1 gives characteristics of sensory nerve fibre types.

Animal models used to investigate the consequences of minor peripheral nerve injury include chronic constriction injury (CCI) where four loose ligatures are applied to the



Table 15.1

Characteristics of sensory nerve fibre types
Morphological fibre type Large myelinated fibres Respond to Different types of mechanoreceptor responding to innocuous stimuli, e.g.:
• •

Pacinian corpuscle - vibration Meissner's corpuscle - light touch Mechanoreceptors Cold-sensitive thermoreceptors Nociceptors responding to noxious pressure and heat Warmth-sensitive thermoreceptors Nociceptor responding to noxious pressure, heat and irritant chemicals

Small my